EP1030047A2

EP1030047A2 - Fuel pressure control device and method for high pressure fuel injection system

Info

Publication number: EP1030047A2
Application number: EP00100737A
Authority: EP
Inventors: Tatsumasa Sugiyama; Yuuichirou Katou
Original assignee: Toyota Motor Corp
Current assignee: Toyota Motor Corp
Priority date: 1999-02-15
Filing date: 2000-01-14
Publication date: 2000-08-23
Anticipated expiration: 2020-01-14
Also published as: EP1030047A3; DE60045229D1; EP1030047B1

Abstract

A fuel pressure control device and method for a high pressure fuel injection system serve to ensure good control and response characteristics irrespective of a transition state or a stationary state. Injectors (12) injecting a fuel into combustion chambers of cylinders (#1 to #4) of an engine (10) are connected to a common rail (20) and supplied with a high pressure fuel from the common rail (20). A fuel pump (30) pressurizes fuel in a fuel tank (14) to high pressure and force-feeds to the common rail (20). An ECU (60) sets closing timings of regulation valves (70a, 70b) of a fuel pump (30) based on predetermined control gains, thereby changing fuel force feed quantity and controlling fuel pressure within the common rail (20) to coincide with target fuel pressure. The ECU (60) variably sets the control gains based on the bulk modulus of the fuel calculated based on the pressure and temperature of the fuel within the common rail (20).

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention

The present invention relates to a fuel pressure control device and method for a high pressure fuel injection system for controlling fuel pressure within an accumulator piping to follow up the change of target fuel pressure by setting the control input of a fuel pump forcedly feeding a high pressure fuel to the accumulator piping to which fuel injection valves are connected.

2. Description of the Related Art

Normally, in an internal combustion engine provided with an accumulator piping such as a common rail, a high pressure fuel is forcedly fed from a fuel pump to the accumulator piping and injected into the combustion chamber of the engine. In this case, the fuel pressure within the accumulator piping, i.e., the fuel injection pressure of the fuel injection valve is controlled to follow up target fuel pressure in accordance with an engine operation state by setting the control input of the fuel pump based on a predetermined control gain. If such fuel pressure control is conducted, the atomized particle size of the injected fuel and the like are adapted to the engine combustion state. The control gain is to determine follow-up speed when the fuel pressure within the accumulator piping follows up the change of the target fuel pressure and the control gain is empirically set in order to ensure predetermined control response characteristics.

Meanwhile, as the pressure, temperature and the like of the fuel forcedly fed by the fuel pump are changed, the change speed of the fuel pressure with respect to the predetermined control input differs even if the fuel pump is operated at fixed operation speed (which corresponds to engine rotation speed for a type of the pump driven by an engine output shaft). If the change speed of the fuel pressure, temperature and the like varies greatly, the follow-up speed of the fuel pressure goes beyond an optimum speed range to possibly cause overshoot phenomenon in which the fuel pressure exceeds target fuel pressure or, conversely, undershoot phenomenon in which the fuel pressure requires more than time necessary to reach the target pressure.

To avoid the disadvantage, as described in, for example, Japanese Patent Application Laid-Open No. 5-106495, follow-up speed is detected at a transient time at which fuel pressure shows great fluctuation and a control gain is updated according to the detected follow-up speed. That is to say, if it is judged that the follow-up speed of the fuel pressure is too high, the control gain is decreased by a predetermined quantity. Conversely, if it is judged that the follow-up speed is too low, the control gain is increased by a predetermined quantity. By repeatedly updating the control gain, it is possible to gradually correct the follow-up speed to fall within an optimum speed range even if the follow-up speed of the fuel pressure goes beyond the optimum speed range.

Nevertheless, the update of the control gain stated above is not initiated until the follow-up speed of the fuel pressure goes beyond the optimum speed range. Besides, the update process is gradual, so that it takes some time to return the follow-up speed to fall within the optimum speed range. Therefore, if the follow-up speed of the fuel pressure greatly goes beyond the optimum speed range in quite short period of time in a transition state, the correction of the follow-up speed by updating the control gain cannot catch up. Thus, it is unavoidable that the control response characteristics become unstable.

Furthermore, even in a stationary state in which the target fuel pressure and fuel pressure within, for example, the accumulator piping are maintained almost constant, the change speed of the fuel pressure with respect to a predetermined control input becomes excessive depending on the magnitude of the fuel pressure or temperature and excess correction operation for correcting the fuel pressure in response to a slight change of the target fuel pressure is performed, thereby inducing a so-called hunting phenomenon. The above-stated control method, however, cannot handle such a hunting phenomenon, with the result that control response characteristics becomes unstable as in the case of the transition state.

SUMMARY OF THE INVENTION

The present invention has been made under these circumstances. It is, therefore, an object of the present invention to provide a fuel pressure control device and method for a high pressure fuel injection system capable of constantly ensuring good control response characteristics in a transition state, a stationary state, or whichever.

To attain this object, a fuel pressure control device for a high pressure fuel injection system according to the present invention is characterized by comprising control means for setting a control input of a fuel pump forcedly feeding a high pressure fuel to an accumulator piping, to which a fuel injection valve is connected, based on a predetermined control gain and thereby controlling fuel pressure within the accumulator piping so as to follow up a change of target fuel pressure; variable element detection means for detecting a variable element making a change speed of the fuel pressure within the accumulator piping variable with respect to the control input at a predetermined operation speed of the fuel pump; and control gain setting means for variably setting the control gain based on the variable element so as to cancel a variation of the change speed of the fuel pressure caused by the variable element in advance.

With the above-stated structure, even if the change speed of the fuel pressure with respect to a predetermined control input is varied by the variable element, the control gain is set based on the variable element and the variation of the change speed can be thereby canceled in advance. Thus, whether in a transition state or in a stationary state, the fuel pressure within the accumulator piping is allowed to follow up every change of the target fuel pressure at appropriate follow-up speed and good control and response characteristics can be, therefore, ensured.

The object of the invention will also be solved by the method according to claim 15.

Although this summary does not describe all the features of the present invention, it should be understood that any combination of the features stated in the dependent claims is within the scope of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic configuration view showing a high pressure fuel injection system;

FIG. 2 is a flow chart showing fuel pressure control procedures in a first embodiment according to the present invention;

FIG. 3 is a timing chart showing the execution ranges of feed forward control and feedback control with respect to target fuel pressure;

FIG. 4 is a graph showing the relationship among fuel pressure, fuel temperature and bulk modulus of elasticity;

FIG. 5 is a flow chart showing procedures for calculating a feedback control correction term;

FIG. 6 is a graph showing the relationship between a final valve closing timing and fuel pump suction speed;

FIG. 7 is a flow chart showing procedures for calculating a feed forward control correction term;

FIG. 8 is a flow chart showing procedures for calculating a control gain in fuel pressure feedback control;

FIG. 9 is a flow chart showing procedures for calculating a control gain in fuel pressure feed forward control;

FIG. 10 is a graph showing the relationship between the volume of a force feed system, the control gain of feedback control and feed forward control;

FIG. 11 shows a cross-sectional view of supply pumps and fuel passages within the fuel pump;

FIG. 12 is a timing chart showing the change states of rail pressure and the like in a stationary state;

FIG. 13 is a flow chart showing fuel pressure control procedures;

FIG. 14 is a flow chart showing fuel pressure control procedures;

FIG. 15 is a timing chart showing the change state of rail pressure in a stationary state;

FIG. 16 is a flow chart showing procedures for calculating a proportional term in fuel pressure control; and

FIG. 17 is a flow chart showing the abnormality detection means of the fuel force feed system.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Now, description will be given to various embodiments in which the present invention is applied to a control device for controlling fuel pressure within a common rail provided in a diesel engine.

[First Embodiment]

FIG. 1 is a schematic configuration view of a four-cylinder direct injection diesel engine (to be referred to simply as an

engine" hereinafter) 10 and a high pressure fuel injection system of the engine 10.

The high pressure fuel injection system is provided with injector 12 disposed corresponding to cylinders #1 to #4 of the engine 10, respectively, a common rail 20 to which the injectors 12 are connected, a fuel pump 30 forcedly feeding a fuel within a fuel tank 14 to the common rail 20 and an electronic control unit (to be referred to as ECU" hereinafter) 60.

The common rail 20 functions to accumulate the fuel supplied from the fuel pump 30 at a predetermined pressure and the fuel injection pressure of the injectors 12 is determined based on the fuel pressure (rail pressure) within the common rail 20. A relief valve 22 is attached to the common rail 20. The relief valve 22 is connected to the fuel tank 14 through a relief passage 21.

The relief valve 22 is normally held in a closed state. When rail pressure increases to exceed a preset upper limit due to some abnormality, it is turned into an open state to forcedly decrease the rail pressure. It is noted that if the rail pressure is greatly decreased as stated above, the ECU 60 judges that fuel leakage occurs to the fuel force feed system, stops fuel injection and forcedly stops the operation of the engine 10.

The injectors 12 are solenoid valves opened and closed by the ECU 60 and inject the fuel supplied from the common rail 20 to the combustion chambers (not shown) of the cylinders #1 to #4, respectively. Each of the injectors 12 is also connected to the fuel tank 14 through the relief passage 21. The fuel leaking into the injectors 12 is returned to the fuel tank 14 through the relief passage 21.

The ECU 60 executes control over the fuel force feed quantity of the fuel pump 30, the fuel injection timing of the injectors 12 and fuel injection quantity and consists of a memory 64 storing various control programs, function data and the like, a CPU 62 executing various arithmetic processing and the like.

Further, various types of sensors for detecting the operation state of the engine 10, a fuel state within the common rail 20 and the like are connected to the ECU 60 and detection signals are inputted from the respective sensors into the ECU 60.

For example, a rotation speed sensor 65 and a cylinder discrimination sensor 66 are provided in the vicinity of a crank shaft (not shown) of the engine 10 and in the vicinity of a cam shaft (not shown), respectively. The ECU 60 calculates the rotation speed (engine rotation speed NE) and rotation angle (crank angle CA) of the crank shaft based on the detection signals inputted from the

sensors

65 and 66, respectively.

An acceleration sensor 67 is provided in the vicinity of an accelerator (not shown) and outputs a detection signal in accordance with the actuating quantity (accelerator opening ACCP) of the accelerator. The common rail 20 is provided with a fuel sensor 68 from which a detection signal is outputted in accordance with rail pressure (detected fuel pressure PCR). A fuel temperature sensor 69 is provided in the middle of the relief passage 21 and outputs a detection signal in accordance with fuel temperature THF. The ECU 60 detects the accelerator opening ACCP, the detected fuel pressure PCR and the fuel temperature THF based on the detection signals from the sensors 67 to 69, respectively.

The fuel pump 30 is provided with a drive shaft 40 rotated by the crank shaft of the engine 10, a feed pump 31 actuated based on the rotation of the drive shaft 40 and a pair of supply pumps (first supply pump 50a and second supply pump 50b) driven by an annular cam 42 formed at the drive shaft 40.

The feed pump 31 sucks a fuel within the fuel tank 14 from a suction port 34 through a suction passage 24 and supplies the fuel to the first supply pump 50a and the second supply pump 50b at predetermined feed pressure, respectively. Among the fuel sucked from the suction port 34, excess fuel supplied to neither the supply pump 50a nor pump 50b is returned to the fuel tank 14 through the relief passage 21 from the relief port 36.

The first supply pump 50a and the second supply pump 50b which are of inner cam type, applies further high pressure (e.g., 25 to 180 MPa) to the fuel supplied from the feed pump 31 based on the reciprocating motion of plungers (not shown) and forcedly feeds the pressurized fuel to the common rail 20 through a discharge passage 23 from a discharge port 38. This fuel force feed operation is conducted by the supply pumps 50a and 50b alternately and intermittently. The fuel pump 30 is provided with the first regulation valve 70a and the second regulation valve 70b for regulating the fuel force feed quantities of the supply pumps 50a and 50b, respectively. Each of the

regulation valves

70a and 70b consists of a solenoid valve and the opening/closing thereof is driven by the ECU 60.

The first regulation valve 70a is opened during the suction stroke of the first supply pump 50a to initiate fuel suction operation and also closed during the suction stroke to stop fuel suction operation. The entire fuel thus sucked is pressurized during a force feed stroke following the suction stroke and forcedly fed to the common rail 20 from the first supply pump 50a. Thus, by changing the opening timing of the first regulation valve 70a, the fuel force feed quantity of the first supply pump 50a is regulated. Likewise, by changing the opening timing of the second regulation valve 70b during its suction stroke, the fuel force feed quantity of the second supply pump 50b is regulated.

The ECU 60 sets the opening timings of the

regulation valves

70a and 70b at those in which respective suction strokes initiate and changes only the closing timings during the respective suction strokes, thereby controlling the fuel force feed quantities of the supply pumps 50a and 50b, respectively. In other words, the ECU 60 increases the fuel force feed quantities of the supply pumps 50a and 50b by delaying the opening timings of the

regulation valves

70a and 70b and thereby elongating their opening timings, and decreases the fuel feed quantities of the supply pumps 50a and 50b by advancing the closing timings and thereby reducing the opening timings of the

regulation valves

70a and 70b.

The opening and closing timings of the

regulation valves

70a and 70b are defined with crank angle as a unit. The closing timing is set as a relative crank angle based on the opening timing. Accordingly, if, for example, fuel force feed operations of the supply pumps 50a and 50b are to be stopped, the closing timings thereof are set at 0° CA" and the

regulation valves

70a and 70b are maintained closed during their suction strokes. On the other hand, if the fuel force feed quantities are set to a maximum, opening timings are set at 180° CA" corresponding to the crank angle variation during the suction strokes and the

respective regulation valves

70a and 70b are constantly kept opened.

The ECU 60 changes the fuel force feed quantity of the fuel pump 30, thereby controlling rail pressure to be coincident with predetermined target fuel pressure. Now, description will be given to how the ECU 60 controls rail pressure hereinafter.

The flow chart of FIG. 2 shows processing procedures for controlling fuel pressure (rail pressure) in this embodiment. The ECU 60 executes this processing routine as an interruption processing for every predetermined crank angle.

First, in step 100, the ECU 60 calculates a basic opening timing ANGBASE based on the engine rotation speed NE and the fuel injection quantity. This basic opening timing ANGBASE is determined based on the necessary fuel force feed quantity of the fuel pump 30 during the stationary operation of the fuel pump 30, i.e., when the rail pressure is almost equal to target pressure (target fuel pressure PCRTRG) and the target fuel pressure PCRTRG is maintained almost constant. The fuel injection quantity is a value calculated based on the accelerator opening ACCP and the engine rotation speed NE in a routine other than the above routine and stored in the memory 64.

The relationship among the basic opening timing ANGBASE, fuel injection quantity and engine rotation speed NE is stored in advance in the memory 64 of the ECU 60 as function data based on characteristics and the like of the engine 10 and of the fuel pump 30. The ECU 60 calculates the basic closing timing ANGBASE by referring to this function data.

Next, in steps 200 and 300, the ECU 60 calculates a feedback (F/B) control correction term ANGFB and a feed forward (F/F) control correction term ANGFF. The F/B control correction term ANGFB and the F/F control correction term ANGFF are set based on the deviation between the target fuel pressure PCRTRG and the rail pressure and on the change speed of the target fuel pressure PCRTRG, respectively. Detailed procedures for calculating the control correction terms AFGFB and ANGFF will be described later.

In step 350, it is determined which to select as a fuel pressure control method, F/B control or F/F control based on the next judgment formula (1): α × ANGFF > ß × ANGFB

In the judgement formula (1), α" and β" are weighting coefficients to determine which to emphasize and adopt as a control method during the transition operation of the fuel pump 30, F/B control or F/F control. In fuel pressure control in this embodiment, as shown in, for example, FIG. 3, F/F control is first executed during the transition operation of the fuel pump 30 during which the target fuel pressure PCRTRG suddenly increases. When the change speed of the target fuel pressure PCRTRG decreases and the operation state of the fuel pump 30 almost turns into a stationary state, F/F control is switched to F/B control. Here, as the weighting coefficient α by which the F/F control correction term ANGFF is multiplied is set smaller relative to the weighting coefficient β by which the F/B control correction term ANGFB is multiplied, F/B control is executed in an earlier stage when operation is switched from the transition operation to the stationary operation.

In step 350, if it is judged that the above judgment formula (1) is satisfied, the ECU 60 goes to step 355 in which the final closing timing ANGFIN is calculated based on the following formula (2) so as to execute F/F control: ANGFIN = ANGBASE + ANGFF

Meanwhile, if it is judged that the judgment formula (1) is not satisfied in step 350, the ECU 60 goes to step 360 in which the final closing timing ANGFIN is calculated based on the following formula (3) so as to execute F/B control: ANGFIN = ANGBASE + ANGFB

After the final closing timing ANGFIN is calculated as described above, the final closing timing ANGFIN is compared with its upper limit 180° CA" and its lower limit 0° CA" in steps 365 to 390 and corrected if necessary. That is, if the final closing timing ANGFIN exceeds 180° CA" ( YES" in step 365), then it is set equal to 180° CA" (in step 370) and if it is below 0° CA" ( YES" in step 380), then it is set equal to 0° CA" (in step 390). Thereafter, the ECU 60 temporarily ends this processing routine.

The ECU 60 outputs driving signals, generated based on the above final closing timing ANGFIN, alternately to the

regulation valves

70a and 70b in a processing routine other than the above routine, thereby regulating the fuel force feed quantity of the fuel pump 30 and controlling the rail pressure. As a result, as indicated by, for example, a two-dot chain line in FIG. 3, the rail pressure is gradually converged into the target fuel pressure PCRTRG while being varied with the fuel force feed operation of the fuel pump 30 and the fuel injection operations of the injectors 12.

Next, description will be given to procedures for calculating the above F/B control correction term ANGFB with reference to the flow chart of FIG. 5.

First, in step 202, the ECU 60 calculates target fuel pressure PCRTRG based on the fuel injection quantity, the engine rotation speed NE and the like. The relationship among the target fuel pressure PCRTRG, the fuel injection quantity and the engine rotation speed NE is experimentally obtained in advance so that the atomized particle size of the fuel injected from the injectors 12 and the like are adapted to the engine combustion state and stored in the memory 64 of the ECU 60 as function data. The ECU 60 calculates the target fuel pressure PCRTRG with reference to this function data.

Next, in step 204, the deviation (PCRTRG - PCR) between the target fuel pressure PCRTFG and detected fuel pressure PCR is compared with predetermined determination pressure P1. The value of the detected fuel pressure PCR is that detected periodically based on the detection signal of the fuel pressure sensor 68 in another processing routine and stored in the memory 64 of the ECU 60. In this processing routine, the value of the detected fuel pressure PCR is read out from the memory 64 at need.

As stated above, the fuel feed operation of the fuel pump 30 is intermittently or alternately executed by the supply pumps 50a and 50b. Due to this, even if the final closing timing ANGFIN is calculated based on the detected fuel pressure PCR and the like, there exists a certain delay until the calculation result is reflected as the change of the fuel force feed quantity of the fuel pump 30. Consequently, if the deviation (PCRTFG - PCR) is large, therefore, the rail pressure suddenly increases, then the variation of the rail pressure during this delay period necessarily increases. The determination pressure P1 is used to judge whether or not the variation of the rail pressure during the delay period is measurably large.

In step 204, if it is judged that the deviation (PCRTRG - PCR) is larger than the determination pressure P1 and the variation of the rail pressure is measurable, processing in steps 206 and 208 are executed to correct the detected fuel pressure PCR based on the variation.

First, in step 206, the bulk modulus E of the fuel is calculated based on the detected fuel pressure PCR and the fuel temperature THF. The memory 64 of the ECU 60 stores the relationship among the bulk modulus E, the detected fuel pressure PCR and the fuel temperature THF as function data. FIG. 4 shows this function data as a three-dimensional map. The bulk modulus E tends to be higher as the detected fuel pressure PCR is higher and the fuel temperature THF is lower. The ECU 60 refers to the function data when calculating the bulk modulus E.

Next, in step 208, predicted fuel pressure PCRPRE is calculated in accordance with the following procedures.

First, in a period (to be referred to as change prediction period" hereinafter) from the detection timing of the detected fuel pressure PCR until the start of the fuel force feed operation reflected by the calculation result of the final closing timing ANGFIN.the ECU 60 calculates:

(a) the quantity of the fuel Q_a forcedly fed to the common rail 20 from the fuel pump 30;

(b) the quantity of the fuel Q_b injected from the injectors 12; and

(c) the quantity of the fuel Q_c leaking from the injectors 12 and returned to the fuel tank 14 through the relief passage 21. For reference, the fuel quantity Q_a in item (a) above can be calculated based on the final closing timing ANGFIN calculated in the previous control period and the like, fuel quantity Q_b in item (b) above can be made equal to the fuel injection quantity and the fuel quantity Q_c in item (C) above can be calculated based on the leaking characteristics of the injectors 12, fuel temperature THF, detected fuel pressure PCR, engine rotation speed NE and the like.

Next, the ECU 60 calculates the variation ΔP of the rail pressure during the change prediction period based on the following formula (4): ΔP = E × (Qa + Qb + Qc)/VCR

In the formula (4), VCR is the total volume of the common rail 20 and discharge passage 23 (to be referred to as force feed system volume" hereinafter) and set at a constant value in this embodiment.

The ECU 60 adds the variation of the rail pressure ΔP thus calculated to the detected fuel pressure PCR and sets the addition result as prediction fuel pressure PCRPRE (= PCR + ΔP).

Next, in step 210, the ECU 60 sets arithmetic fuel pressure TPCRFB used in various arithmetic operations for F/B control. The arithmetic fuel pressure TPCRFB corresponds to the rail pressure after the change prediction period. In step 210, since it is judged that the rail pressure changes greatly in the change prediction period, the predicted fuel pressure PCRPRE is set as the arithmetic fuel pressure TPCRFB. On the other hand, in step 204, if it is judged that the deviation (PCRTRG - PCR) is not greater than the determination pressure P1 and the variation of the rail pressure can be negligible, then the detected fuel pressure PCR is set as arithmetic fuel pressure TPCRFB in step 211.

In step 212, the arithmetic fuel pressure TPCRFB thus set is subtracted from the target fuel pressure PCRTRG and the subtraction result (PCRTRG - TPCRFB) is set as a deviation ΔPCRFB.

Next, in step 214, the ECU 60 calculates bulk modulus EFB used to calculate the control gain KFB of F/B control based on the arithmetic fuel pressure TPCRFB and fuel temperature THF. Here, the ECU 60 refers to the function data shown in FIG. 4 as in the case of the processing in step 206 and thereby calculates this bulk modulus EFB.

In step 215, the ECU 60 calculates the control gain KFB based on the following formula (5): KFB = VCR / (ΔVPUMP × EFB)

In the formula (5), ΔVPUMP is the variation of the suction quantity of the fuel pump 30 occurring when the final opening timing ANGFIN is changed by a unit angle (1° CA) and, in the fuel pump 30 in this embodiment, the variation of the suction quantity is equal to that of the force feed quantity since all the sucked fuel during the suction stroke is forcedly fed during the force feed stroke.

In addition, the fuel suction and force feed operations of the fuel pump 30 are conducted by the reciprocating motion of the plunger following the rotation of the cam 42. The reciprocation speed of this plunger is changed according to the rotation phase of the cam 42, i.e., crank angle CA. Thus, the variation of the force feed quantity (suction quantity variation) ΔVPUMP also differs according to the magnitude of the final closing timing ANGFIN.

The suction speed of the fuel pump 30 and the final closing timing ANGFIN have a relationship as shown in, for example, FIG. 6 and the suction speed differs according to the magnitude of the final closing timing ANGFIN. Due to this, as shown therein, if the values of the final closing timing ANGFIN (ANGFIN1, ANGFIN2) differ from each other, the values of the suction quantity variation (areas in the oblique-line regions in FIG. 6) corresponding to the respective values ANGFIN1 and ANGFIN2, i.e., the values of the variation of the force feed quantity ΔVPUMP (ΔVPUMP1, VPUMP2) differ as well.

Thus, in this processing routine, the average suction speed (indicated by the two-dot chain line in FIG. 6) during the suction stroke (180° CA) of the fuel pump 30 is calculated and the variation of the suction quantity which occurs when the final closing timing ANGFIN is changed by the unit angle while the suction speed is made constant and equal to the average speed is set as the variation of the force feed quantity ΔVPUMP.

After calculating the control gain KFB as stated above, the F/B control correction term ANGFB is calculated based on the following formula (6) in step 218: ANGFB = KFB × ΔPCRFB

Meanwhile, if the bulk modulus increases according to the change of the fuel pressure or temperature, the change speed of the rail pressure with respect to the same fuel force feed quantity increases and the follow-up speed of the rail pressure with respect to the change of the target fuel pressure PCRTRG increases.

In this respect, according to the F/B control in this embodiment, the F/B control correction term ANGFB is calculated based on the control gain KFB reset to be relatively small as is obvious from the above formulas (5) and (6) and the increase of the follow-up speed is therefore avoided in advance. Conversely, if the bulk modulus EFB decreases, the F/B control correction term ANGFB is calculated based on the control gain KFB reset to be relatively large and the decrease of the follow-up speed can be therefore avoided in advance.

After the F/B control correction term ANGFB is thus calculated, the ECU 60 calculates an F/F control correction term ANGFF. FIG. 7 is a flow chart showing procedures for calculating the F/F control correction term ANGFF.

First, the ECU 60 compares target fuel pressure PCRTRG and detected fuel pressure PCR in step 302. Here, if it is judged that the detected fuel pressure PCR is not less than the target fuel pressure PCRTRG, the F/F control correction term ANGFF is set at 0° CA" in step 315 and the processing in this routine is temporarily finished.

On the other hand, if it is judged that the target fuel pressure PCRTRG is higher than the detected fuel pressure PCR in step 302, the ECU 60 moves to a processing in step 304. In step 304, the target fuel pressure PCRTRG is subtracted from the previous target fuel pressure value PCRTRG0, the absolute value |PCRTRG - PCRTRG0| is compared with predetermined determination pressure P2 and it is thereby judged whether or not the target fuel pressure PCRTRG varies greatly, that is, whether or not the change of the target fuel pressure PCRTRG is in a transition state.

If it is judged that the variation of the target fuel pressure PCRTRG is small (in an almost stationary change state), arithmetic fuel pressure TRCRFF is set equal to the target fuel pressure PCRTRG in step 306. This arithmetic fuel pressure TPCRFF is used for various arithmetic operations for F/F control and, as will be described later, the F/F control correction term ANGFF is calculated based on the deviation (PCRTRG - TPCRFF) between the arithmetic fuel pressure TPCRFF and the target fuel pressure PCRTRG.

Meanwhile, if it is judged that the variation of the target fuel pressure PCRTRG is large and the change of the target fuel pressure PCRTRG is in a transition state in step 304, then the subtraction value (PCRTRG - PCRTRG0) is compared with the determination pressure P2 in step 305 and it is thereby judged whether the target fuel pressure PCRTRG is on the increase or on the decrease. If it is judged that the target fuel pressure PCRTRG is on the decrease, the arithmetic fuel pressure TPCRFF is set equal to the target fuel pressure PCRTRG0 in step 308.

On the other hand, if it is judged that the target fuel pressure PCRTRG is on the increase in step 305, the previous target fuel pressure PCRTRG0 is compared with the detected fuel pressure PCR and the higher pressure is set as arithmetic fuel pressure TPCRFF in step 307.

In step 307, the deterioration of control characteristics is suppressed. While the target fuel pressure PCRTRG is on the decrease and the rail pressure is higher than the target pressure PCRTRG, if the target fuel pressure PCRTRG starts increasing, the relationship (PCRTRG > PCR > PCRTRG0) may be created, though temporarily, among the magnitudes of the target fuel pressure PCRTRG, the detected fuel pressure PCR and of the previous target fuel pressure PCRTRRG0.

Even in this case, as in the case of step 308, if the arithmetic fuel pressure TPCRFF is set equal to the previous target fuel pressure PCRTRG0, the deviation (PCRTRG - TPCRFF) increases by the difference between the detected fuel pressure PCR and the previous target fuel pressure PCRTRG0. Due to this, the F/F control correction term ANGFF is set at an excessive high value and excess fuel is forcedly fed.

In that case, therefore, if the arithmetic fuel pressure TPCRFF is set, it is set not equal to the previous target fuel pressure PCRTRG0 but the detected fuel pressure PCR thereby setting the F/F control correction term ANGFF to be an appropriate value.

After the arithmetic fuel pressure TPCRFF is calculated in the respective steps 306 to 308, bulk modulus EFF used to calculate a control gain KFF for F/F control is calculated based on the arithmetic fuel pressure TPCRFF and the fuel temperature THF in step 310. Here, the ECU 60 refers to the function data shown in FIG. 4 as in the case of steps 206 and 214 and thereby calculates the bulk modulus EFF.

Then, the ECU 60 calculates the control gain KFF in step 311 based on the following formula (7): KFF = VCR / (ΔVPUMP × EFF)

Next, in step 314, the F/F control correction term ANGFF is calculated based on the following formula (8): ANGFF = ANGFF0 + KFF × ΔPCRFF

In the formula (8), ΔPCRFF is the deviation (PCRTRG - TPCRFF) between the target fuel pressure PCRTRG and the arithmetic fuel pressure TPCRFF. ANGFF0 is the carry-over quantity of the F/F control correction term ANGFF and calculated in the next steps 316 and 318.

In step 316, the addition value (ANGBASE + ANGFF) of the basic closing timing ANGBASE and the F/F control correction term ANGFF calculated this time is compared with the set upper limit 180° CA" of the addition value.

If the addition value (ANGBASE + ANGFF) exceeds the upper limit (180° CA), the difference (ANGBASE + ANGFF - 180° CA) between the addition value (ANGBASE + ANGFF) and the upper limit (180° CA) is added to the present carry-over quantity ANGFF0 and the resultant addition value (ANGFF0 + (ANGBASE + ANGFF - 180° CA)) is set as a new carry-over quantity ANGFF0 in step 318.

On the other hand, if the addition value (ANGBASE + ANGFF) is not more than the upper limit (180° CA), the carry-over quantity ANGFF0 is not updated, that is, the processing in step 318 is skipped.

It is noted here that, as is obvious from the formulas (7) and (8), as the bulk modulus EFF is higher, the control gain KFF is set smaller in F/F control in this embodiment like F/B control. Due to this, even if the bulk modulus EFF is changed, an F/F control correction term ANGFF is calculated as a value capable of canceling, in advance, the variation of the change speed of the rail pressure resulting from the change of the bulk modulus EFF.

After the processing in step 318 is executed or if it is judged that the update of the carry-over quantity ANGFF0 is not necessary in step 316, the ECU 60 sequentially executes processing in step 350 and the following shown in FIG. 2.

As described above, according to the fuel pressure control in this embodiment, the control gains (KFB, KFF) for F/B control and F/F control, respectively, are not constant but reset at intervals at which the control input of the fuel pump 30 (final closing timing ANGFIN) is calculated. If these control gains are set, the magnitudes of the bulk moduli (EFB, EFF) of the fuel are constantly monitored and the magnitudes are reflected in control gain setting. By doing so, the variation of the change speed of the rail pressure relative to the control input (final closing timing ANGFIN) of the fuel pump 30 resulting from the change of the bulk moduli is canceled in advance.

According to this embodiment, therefore, it is possible to maintain the follow-up speed of the rail pressure to be appropriate speed irrespectively of the changes of the bulk moduli. Thus, in a transition state in which the target fuel pressure greatly varies, it is possible to ensure preventing the occurrence of overshoot phenomenon or undershoot phenomenon, whereas in a stationary state, it is possible to ensure preventing the occurrence of hunting phenomenon. As a result, it is possible to ensure good control response characteristics whether in the transition state or in the stationary state.

Furthermore, since the bulk moduli are calculated based on both the fuel pressure and the fuel temperature, they can be accurately detected based on parameters which can be obtained relatively easily and it can be ensured reflecting their variations in control gain setting. Thus, it is possible to ensure good control response characteristics more surely.

[Second Embodiment]

Next, the second embodiment of the present invention will be described, while centering around the difference between the first and second embodiments. It is noted that the same constituent elements as those in the first embodiment will be not described herein.

In fuel pressure control in this embodiment, when the abnormality of the fuel temperature sensor 69 is detected, procedures for setting a control gain KFB for F/B control and a control gain KFF for F/F control are changed.

Here, the abnormality of the fuel temperature sensor 69 means the change of an output signal due to disconnection of connection wiring, the deterioration of sensor elements or the like. Such abnormality detection is made based on the judgement as to whether or not the output signal falls within a predetermined output range in another abnormality judgment processing executed by the ECU 60. Also, if the fuel temperature 69 is judged abnormal in this abnormality determination processing, then an abnormality detection flag XTHF indicating occurrence of abnormality is set ON".

If abnormality like this occurs to the fuel temperature sensor 69, the control accuracy of various controls such as fuel injection control and fuel pressure control with the fuel temperature THF used as a control variable, deteriorates. In this case, unlike occurrence of abnormality such as a case where the injectors 12 remain open or stuck, it is less necessary to urgently stop the engine 10.

However, if abnormality occurs to the fuel temperature sensor 69 and thereby an actual fuel temperature differs from the fuel temperature THF detected by the fuel temperature sensor 69, each bulk modulus (EFB, EEF) is set lower than the value corresponding to the actual fuel temperature and the control input of the fuel pump 30 (final closing timing ANGFIN) is calculated based on the excessive control gains (KFB, KFF).

If the rail pressure is controlled based on such excessive control gains (KFB, KFF) and the target fuel pressure suddenly increases to be close to the upper limit of the rail pressure by, for example, the sudden acceleration of the engine 10, the rail pressure may exceed an upper limit and further increases (overshoots) and the relief valve 22 may be opened. As a result, the operation of the engine 10 may be forcedly stopped.

Further, if the target fuel pressure suddenly decreases to be close to the lower limit by, for example, the sudden deceleration of the engine 10, then the rail pressure decreases (overshoots) to go further below the lower limit, the exhaust property may be deteriorated due to the fuel injection at extremely low fuel injection pressure or the engine 10 may be stopped due to inability to conduct normal engine combustion at worst.

In fuel pressure control in this embodiment, therefore, during the abnormality of the fuel temperature sensor 69, secondary abnormality following the abnormality of the fuel temperature sensor 69, such as occurrence of unintended engine stall or the deterioration of the exhaust property as stated above, is prevented from occurring as much as possible by changing the control gains KFB and KFF to values corresponding to the abnormality. Now, this fuel pressure control will be described with reference to FIGS. 8 and 9. It is noted that the FIGS. 8 and 9 only show the changed parts in the processing of the fuel pressure control shown in the flow charts of FIGS. 2, 5 and 7.

As shown in FIG. 8, the ECU 60 calculates the control gain KFB for F/B control in step 215 and then moves to a processing in step 216. In step 216, it is judged whether or not the abnormality detection flag XTHF is ON".

If it is judged that the abnormality detection flag XTHF is ON" and abnormality occurs to the fuel temperature sensor 69, then the ECU 60 resets the control gain KFB in step 217 in accordance with the following procedures.

First, the ECU 60 refers to the function data shown in FIG. 4 and thereby calculates the bulk modulus EFB as in the case of the processing in step 214. It is noted, however, that the ECU 60 calculates the bulk modulus EFB with the fuel temperature THF set as the lowest temperature THFLOW on the function data in step 217. The calculated bulk modulus EFB is, therefore, always the highest value among those with respect to the same arithmetic fuel pressure TPCRFB.

Next, the ECU 60 calculates the control gain KFB based on the above formula (5) as in the case of the processing in step 215. Here, since the bulk modulus EFB is set relatively high as stated above, the control gain KFB is always calculated to be relatively low in step 217 with respect to the value calculated in step 215.

If it is judged that the fuel temperature sensor 69 is normal in step 216 or after the processing in step 217 is executed, the ECU 60 executes processing in step 218 and the following.

Next, procedures for setting a control gain KFF for F/F control will be described.

As shown in FIG. 9, the ECU 60 calculates the control gain KFF for F/F control in step 311 and then moves to a processing in step 312. In step 312, it is judged whether or not the abnormality detection flag XTHF is ON".

If it is judged that the abnormality detection flag XTHF is ON" and abnormality occurs to the fuel temperature sensor 69, then the ECU 60 resets the control gain KFF in step 313 in accordance with the following procedures.

First, the ECU 60 refers to the function data shown in FIG. 4 and thereby calculates the bulk modulus EFF again with the fuel temperature THF set at the lowest temperature THFLOW as in the case of the processing in step 217. Here, the calculated bulk modulus EFF is the highest value among those with respect to the arithmetic fuel pressure TPCRFF.

As in the case of the processing in step 311, the control gain KFF is calculated based on the above formula (7). Since the bulk modulus EFF is set relatively high, the control gain KFF is always calculated relative low with respect to the value calculated in step 311.

On the other hand, if it is judged in step 312 that the fuel temperature sensor 69 is normal or after the processing in step 313 is executed, the ECU 60 executes processing in step 314 and the following.

As can be seen from the above, according to this embodiment, when the abnormality of the fuel temperature sensor 69 is detected, the control gains KFB and KFF are reset based on the bulk moduli EFB, EFF and the like with the fuel temperature set at the lowest temperature THFLOW.

Due to this, the control gains KFB and KFF are set at the lowest values compared with those set with respect to the arithmetic fuel pressures TPCRFB and TPCRFF, respectively. The absolute values of the F/B control correction term ANGFB and F/F control correction value ANGFF are also set at relatively low values. Therefore, even if either F/F control or F/B control is selected as a fuel pressure control method, the change speed of the final closing timing ANGFIN is lower than the normal value and the change speed of the rail pressure is suppressed, so that the rail pressure follows up more gradually the change of the target fuel pressure (PCRTRG).

Consequently, in this embodiment, when abnormality occurs to the fuel temperature sensor 69, it is possible to suppress occurrence of overshoot phenomenon and to avoid occurrence of engine stall or deterioration of the exhaust property following occurrence of such overshoot phenomenon and suppress the abnormal state from deteriorating further.

In this embodiment, the procedures for setting the control gains KFB and KFF are changed based on the abnormality of the fuel temperature sensor 69. However, it is also possible to detect the abnormality of another high pressure fuel injection system and to change setting procedures based on the detection result.

Further, when the abnormality of the fuel temperature sensor 69 is detected, the control gains KFB and KFF can be set at constant values. The constant values may involve, for example, the minimum values of the control gains KFB and KFF.

In the above-stated embodiments, when the control gains KFB and KFF are calculated, the total volume of the common rail 20 and the discharge passage 23, i.e., the force feed system volume VCR is set at a constant value. If a structure in which the volume of the common rail 20 and the length of the discharge passage 23 are variable is adopted, the control gains KFB and KFF may be changed based on the magnitude of the force feed system volume VCR.

In the latter case, in steps 215 and 311 shown in FIGS. 5 and 7, the present force feed system volume VCR is detected in the both steps and the detected volume VCR is assigned to the formulas (5) and (7) and then the control gains KFB and KFF are calculated. With this structure, as shown in FIG. 10, as the force feed system volume VCR increases and the increase speed of the rail pressure with respect to the unit force feed quantity, the control gains KFB and KFF are set higher. Therefore, in addition to the bulk moduli (EFB, EFF), the follow-up speed of the rail pressure can be maintained to be appropriate speed despite the change of the force feed system volume VCR.

In the above-stated embodiments, the variation of the suction quantity when changing the final closing timing ANGFIN by a unit angle at the average suction speed during the suction stroke of the fuel pump 30 (180° CA), is set as the above-stated variation of the force feed quantity ΔVPUMP and the variation ΔVPUMP is used to operate the control gains KFB and KFF.

Actually, however, as shown in FIG. 6, the magnitude of the suction speed varies according to the final closing timing ANGFIN. Due to this, if the final closing timing ANGFIN falls within an angle range A2, the variation of the force feed quantity ΔVPUMP is estimated slightly larger than the actual variation and the control gains KFB and KFF of the fuel pump 30 may be set relatively low. On the other hand, if the final closing timing ANGFIN falls within an angle range A4, the variation ΔVPUMP is estimated slightly smaller than the actual variation and the control gains KFB and KFF are set relatively high. Thus, it is preferable to consider the operation characteristics of the fuel pump 30 in order to accurately obtain the control gains KFB and KFF.

Accordingly, it is further preferable that correction coefficients KA1 to KA5 corresponding to errors of the variation ΔVPUMP in the angle ranges A1 to A5 shown in FIG. 6 are set, respectively, in advance, after steps 355 and 360 shown in FIG. 2, the final closing timing ANGFIN is corrected based on the correction coefficient KA_n, (where n = 1 to 5).

Specifically, when the final closing timing ANGFIN is corrected under F/B control, for example, correction term f (KA_n, EFB and ΔVPCRFB) is calculated based on the correction coefficient KA_n (where n = 1 to 5), the bulk modulus EFB and the deviation ΔPCRFB. Then, the resultant correction term f is added to the final closing timing ANGFIN calculated in step 355 and the addition value is reset as a new final closing timing ANGFIN.

Furthermore, when the final closing timing ANGFIN is corrected under F/F control, for example, the correction term f (KA_n, EFF, ΔPCRFF) is calculated based on the correction coefficient KA_n, the bulk modulus EFF, the deviation PCREF and the carry-over quantity ANGFF0. Thereafter, the correction term f is added to the final closing timing ANGFIN calculated in step 360 and the addition value is reset as a new final closing timing ANGFIN.

By making such corrections, the control gains KFB and KFF are corrected substantially based on the difference in the variation of the force feed quantity ΔVPUMP with respect to the final closing timing ANGFIN and the error of the final closing timing ANGFIN due to those of the control gains KFB and KFF can be corrected. As a result, even if the relationship between the force feed quantity variation Δ VPUMP and the final closing timing ANGFIN varies, as the operation characteristics of the fuel pump 30, according to the final closing timing ANGFIN and the follow-up speed of the rail pressure with respect to the target fuel pressure is changed due to the variation, the change can be canceled in advance and the follow-up speed can be maintained to be an appropriate value.

Moreover, with the above-stated structure, the bulk moduli EFB and EFF can be regarded as constant values and the control gains KFB and KFF can be set merely based on the force feed system volume VCR and the force feed quantity variation ΔVPUMP.

In the above-stated embodiments, the bulk moduli (EFB, EFF) of the fuel are calculated based on both the fuel pressure and the fuel temperature. It is also possible to calculate them while either the fuel pressure or the fuel temperature is regarded constant.

Further, the fuel temperature (THF) can be estimated based on parameters changing in correlation with the fuel temperature, e.g., the temperature of cooling water of the engine 10.

In the above-stated embodiments, the discharge passage 23 from the fuel pump 30 to the common rail 20 is common to the supply pumps 50a and 50b. If the discharge passage 23 is provided for each of the supply pumps 50a and 50b, the force feed system volume VCR may be changed for each of the supply pumps 50a and 50b to thereby calculate the control gains (KFB, KFF) for F/B control and F/F control.

Likewise, if the variation of the force feed quantity ΔVPUMP differs between the supply pumps 50a and 50b, the variation ΔVPUMP may be changed for each of the supply pumps 50a and 50b to thereby calculate the control gains (KFB, KFF).

Additionally, while both F/B control and F/F control are executed in the above-stated embodiments, the rail pressure can be controlled only by either F/B control or F/F control.

[Third Embodiment]

FIGS. 11A and 11B show the cross-sectional structure of the supply pumps 50a and 50b taken along line 11-11 of FIG. 1 and the schematic configuration view of fuel passages in the fuel pump 30.

As shown therein, the first supply pump 50a has a cylindrical support 43 formed in a housing 41 (see FIG. 1) of the fuel pump 30, a pair of plungers 54a supported to be movable in a reciprocating manner by a through hole 43a formed in the support 43 and the like, the first pressurizing chamber 52a determined by the inner end faces of the plungers 54a and the inner wall of the through hole 43a, and the like. A shoe 55a is formed at the outer end portion of each of the plungers 54a and a roller 56a is rotatably supported by the shoe 55a.

As for a cam 42, a cam face 42c against which each roller 56a can be abutted has an ellipsoidal cross section. Due to this, if the cam 42 rotates following the rotation of the drive shaft 40, the length La between the cam faces 42c in the reciprocating direction of the plungers 54a is increased and decreased according to the rotation. Therefore, if the cam 42 rotates while the rollers 56a are abutted against the cam faces 42c, respectively, the plungers 54a reciprocate such that they are close to and distant from each other. The volume of the first pressurizing chamber 52a is changed according to the reciprocating motions. It is assumed hereinafter that a period in which the distance La between the cam faces 42 increases is a suction stroke" of the first supply pump 50a and that in which the distance La decreases is a force feed stroke".

The drive shaft 40 has a deceleration ratio to a crank shaft set at 1/2 and rotates once whenever the crank shaft rotates twice. Therefore, while the cylinders #1 to #4 perform one-cycle operation including suction, compression, expansion and exhaust while the crank shaft rotates twice, the suction stroke and the force feed stroke are alternately conducted twice in the first supply pump 50a.

The first pressurizing chamber 52a is connected to the field pump 31 through a non-return valve 44a and the first regulation valve 70a and connected to a discharge port 38 through another non-return valve 46a. The non-return valves 44a and 46a restrict the flow of a fuel from the first pressurizing chamber 52a toward the field pump 31 and that from the discharge port 38 toward the first pressurizing chamber 52a, respectively, so that the fuel flows generally from the field pump 31 toward the common rail 20 through the first supply pump 50a.

The first supply pump 50a, the first regulation valve 70a, the non-return valves 44a and 46a, the common rail 20, the field pump 31 and fuel passages connecting them as stated above constitute the first fuel force feed system as a whole.

In the first fuel force feed system, if the first regulation valve 70a is opened during the suction stroke of the first supply pump 50a, the fuel is supplied into the first pressurizing chamber 52a from the field pump 31 through the non-return valve 44a. The entire fuel thus supplied into the first pressurizing chamber 52a is forcedly fed to the discharge port 38 from the first pressurizing chamber 52a through the non-return valve 46a in the force feed stroke of the first supply pump 50a.

Meanwhile, the second supply pump 50b is also of inner cam type and provided with the second pressurizing chamber 52b, plungers 54b, shoes 55b, rollers 56b and the like as in the case of the first supply pump 50a.

A through hole 43b supporting the plunger 54b so as to allow the plunger 54b to reciprocate is formed to extend in the direction orthogonal to the through hole 43a of the first supply pump 50a. Due to this, if it is assumed that a period in which the length Lb between cam faces 42c in the reciprocating directions of the plungers 54b is a suction stroke" of the second supply pump 50b and that in which the length Lb decreases is a force feed stroke" thereof, the suction stroke and force feed stroke of the second supply pump 50b have phases, as crank angles CA, shifted by 180° CA from the suction stroke and the force feed stroke of the first supply pump 50a, respectively.

The second pressurizing chamber 52b has the same structure as that of the first pressurizing chamber 52a. The chamber 52a is connected to the field pump 31 through a non-return valve 44b and the second regulation valve 70b and also connected to the discharge port 38 through another non-return valve 46b.

The second supply pump 50b, the second regulation valve 70b, the non-return valves 44b and 46b, the common rail 20, the field pump 31 and the fuel passages connecting them as stated above constitute the second fuel force feed system as a whole.

FIG. 12 is a timing chart showing the change mode of the rail pressure in a stationary state in which the target value of the rail pressure (target fuel pressure PCTRG) is maintained almost constant, the fuel force feed and suction operation timing of the supply pumps 50a and 50b with respect to the crank angle CA and the like.

As indicated by symbol (a) in FIG. 11, the rail pressure is constantly changed by the fuel force feed operations ((b), (c)) of the supply pumps 50a and 50b and by the fuel injection operations of the injectors (12) ((d)) even in the stationary state. The reason that the rail pressure slightly decreases even in periods in which neither the fuel force feed operation nor the fuel injection operation is executed is that a small quantity of the fuel from the injectors 12 is constantly returned to the fuel tank 14 through the relief passage 21.

Further, the force feed operations of the supply pumps 50a and 50b and the fuel injection operations of the injectors 12 are executed while maintaining their predetermined relationships with respect to the crank angle CA. For example, in a period in which the crank angle CA is [CA1 to CA2], the fuel injection of the injector 12 corresponding to the first cylinder #1 and the fuel force feed operation of the second supply pump 50b are conducted. In a period in which the crank angle CA is [CA2 to CA3], the fuel injection of the injector 12 corresponding to the third cylinder #3 and the fuel force feed operation of the first supply pump 50a are conducted.

Symbol (e) in FIG. 11 indicates the detection timing at which actual fuel pressure PCR is detected by the ECU 60. The actual fuel pressure PCR is detected at predetermined crank angle intervals (180° CA's intervals). The detection timing is set at timing (CA1, CA2, CA3, CA4) at which the rise of the rail pressure by the fuel force feed operations of the supply pumps 50a and 50b end.

Symbol (f) in FIG. 11 indicates the change mode of the determination counter value CPCYLND. This determination counter value CPCYLND is used to conduct processing for calculating the control inputs of the supply pumps 50a and 50b, i.e., the opening timings of the

regulation valves

70a and 70b and is incremented by one at predetermined crank angle (180° CA) intervals in the mode of [.. → 0 → 1 → 2 → 3 → 0 → ..] in a counter value operation routine.

Further, it is possible to judge to which fuel force feed the actual fuel pressure PCR corresponds, the first supply pump 50a or the second supply pump 50b based on the determination counter value CPCYLND.

That is to say, if the determination counter value CPCYLND is set at 1" or 3" when the actual fuel pressure PCR is detected, it can be judged that the actual fuel pressure PCR corresponds to the rail pressure which has risen by the fuel force feed operation of the first supply pump 50a. If the determination counter value CPCYLND is set at 0" or 2", it can be judged that the actual fuel pressure PCR corresponds to the rail pressure which has risen by the fuel force feed operation of the second supply pump 50b.

Next, description will be given to fuel pressure control in this embodiment. In this fuel pressure control, the closing timings (final closing timing ANGFIN) of the

regulation valves

70a and 70b are calculated and the fuel suction quantities of the supply pumps 50a and 50b are changed based on the final closing timing ANGFIN, thereby feedback-controlling the pressure of the fuel forcedly fed from the supply pumps 50a and 50b to coincide with the target value (target fuel pressure PCRTRG).

Now, procedures for controlling fuel pressure will be described in detail with reference to the flow charts shown in FIGS. 13 and 14. A fuel pressure control routine" shown in these flow charts is executed by the ECU 60 as interruption processing at predetermined crank angle (180° CA) intervals.

It is noted that the detection of actual fuel pressure PCR is conducted as part of processing this fuel pressure control routine". Namely, the interruption timing of the fuel pressure control routine" is set, as the detection timing of the actual fuel pressure PCR, at timing (CA1, CA2, CA3 and CA4 shown in FIG. 12) at which the rise of the rail pressure by the force feed operations of the fuel of the supply pumps 50a and 50b ends as already described above.

When this fuel pressure control routine" starts, the ECU 60 detects actual fuel pressure PCR in step 410 and then calculates target fuel pressure PCRTRG based on the fuel injection quantity and engine rotation speed NE in step 420.

The relationship among the target fuel pressure PCRTRG, the fuel injection quantity and the engine rotation speed NE is experimentally obtained so that the atomized particle size and the like of the injected fuel are adapted to an engine combustion state, and stored in the memory 64 of the ECU 60 as function data for calculating the target fuel pressure PCRTRG.

Further, in a routine other than the above routine, the fuel injection quantity is calculated based on the accelerator opening ACCP, the engine rotation speed NE and the like and stored in the memory 64.

Next, in step 430, the ECU 60 calculates a basic closing timing ANGBASE based on the fuel injection quantity, the actual fuel pressure PCR and the engine rotation speed NE.

In step 440, the ECU 60 subtracts the actual fuel pressure PCR from the target fuel pressure PCRTRG and sets the resultant subtraction value as a deviation ΔPCR( = PCRTRG - PCR). In step 450, the ECU 60 calculates a proportional term ANGPRO based on the following formula (9): ANGPRO = KP × ΔPCR

In the formula (9), the proportional term ANGPRO is to conduct proportional operation in so-called PID control (or particularly PI control in this case) and is a correction angle for correcting the basic closing timing ANGBASE according to the magnitude of the deviation ΔPCR. Symbol K_P is a proportional gain and set based on the average force feed characteristics of the respective fuel force feed systems.

Next, in step 460, the ECU 60 judges whether or not update prohibition conditions for an integral term ANGINT are satisfied. This integral term ANGINT is conducts integral operation in PID control and is a correction angle for correcting the basic closing timing ANGBASE according to the magnitude of the integral value of the deviation ΔPCR.

In step 460, the update prohibition conditions are that the previous value ANGFIN0 of the final closing timing ANGFIN is a maximum 180° CA" or a minimum 0° CA", i.e., the rail pressure is in a transition state. That is to say, the integral term ANGINT is used to cancel the stationary deviation remaining between the rail pressure and the target fuel pressure PCRTRG when the rail pressure becomes almost equal to the target fuel pressure PCRTRG. Due to this, the update of the integral term ANGINT is prohibited in the transition state in which the rail pressure suddenly increases.

If the ECU 60 judges that the above update prohibition conditions are satisfied, it sets the update quantity DANGINT at 0° CA" in step 180. On the other hand, if the ECU 60 judges the update prohibition conditions are not satisfied, it moves to processing in step 170, in which the update quantity DANGINT is calculated based on the following formula (10): DANGINT = K1 × |ΔPCR|

In the formula (10), symbol K1 is an integral gain (integral time) and |ΔPCR| is the absolute value of the deviation ΔPCR.

After setting the update quantity DANGINT in step 470 or 480, the ECU 60 calculates the integral term ANGINT based on the update quantity DANGINT in step 355.

Now, detailed procedures for calculating the integral term ANGINT will be described with reference to the flow chart of FIG. 14.

In this embodiment, at the time of calculating the integral term ANGINT, values ANGINT1 and ANGINT2 of the integral term ANGINT corresponding to the fuel force feed systems, respectively are prepared and these values ANGINT1 and ANGINT2 are individually updated synchronously with the fuel force feed operations of the respective fuel force feed systems and stored in the memory 64.

First, in step 502, the ECU 60 judges whether the above-stated determination counter value CPCYLND is 0" or 2", i.e., whether or not the final closing timing ANGFIN to be calculated this time is intended to operate the second fuel force feed system (the second regulation valve 70b of the second supply pump 50b).

If it is judged that the determination counter value CPCYLND is 0" or 2", the integral term ANGINT is set to correspond to the second fuel force feed system in steps 510 to 514.

First, the ECU 60, in step 210, judges whether or not the deviation ΔPCR is higher than 0", i.e., whether or not the target fuel pressure PCRTRG exceeds the actual fuel pressure PCR.

If it is judged that the target fuel pressure PCRTRG exceeds the actual fuel pressure PCR, the ECU 60 reads out the value ANGINT2 of the integral term ANGINT (integral term ANGINT2) corresponding to the second fuel force feed system from the memory 64, updates a value (ANGINT2 + DANGINT) obtained by adding the update quantity DANGINT to the integral term ANGINT2 as a new integral term ANGINT2 and stores the new integral term ANGINT2 in the memory 64 in step 212.

On the other hand, if it is judged that the target fuel pressure PCRTRG is not more than the actual fuel pressure PCR, the ECU 60 updates a value (ANGINT2 - DANGINT) obtained by subtracting the update quantity DANGINT from the value ANGINT2 of the integral value ANGINT as a new integral term ANGINT2 and stores the new integral term ANGINT2 in the memory 64 in step 213.

If the above-stated update prohibition conditions are satisfied, the update quantity DANGINT is set at 0° CA" and the update of the integral term ANGINT2 in step 212 or 213 is not, therefore, substantially executed.

By executing the processing in step 212 or 213 the integral term ANGINT2 is integrated based on the deviation Δ PCR generated by the fuel force feed operation of the second fuel force feed system.

In step 514, the ECU 60 sets the integral term ANGINT2 thus updated corresponding to the second fuel force feed system as a final integral term ANGINT.

Meanwhile, in the previous step 502, if it is judged that the determination counter value CPCYLND is 1" or 3", the ECU 60 integrates the value ANGINT1 of the integral term ANGINT corresponding to the first fuel force feed system in steps 520 to 524 as in the case of the processing in steps 510 to 514 and sets the resultant value as a final integral term ANGINT.

After setting the integral term ANGINT corresponding to each of the fuel force feed systems as stated above, the ECU 60 judges whether or not the integral term ANGINT finally set and the integral terms ANGINT1 and ANGINT2 of the respective fuel force feed systems stored in the memory 64 fall within a predetermined angle range in step 540.

If it is judged that some of the integral terms ANGINT, ANGINT1 and ANGINT2 exceeds a predetermined upper limit or falls below a predetermined lower limit, the ECU 60 corrects the integral term ANGINT, ANGINT1 or ANGINT2 which is out of the predetermined angle range, to be equal to the upper limit angle or the lower limit angle in step 242.

By limiting the magnitudes of the integral terms ANGINT, ANGINT1 and ANGINT2 to fall within the predetermined angle range as stated above, these terms ANGINT, ANGINT1 and ANGINT2 are temporarily set excess and can suppress the state which does not correspond to the magnitude of the present stationary deviation from being maintained for a long time.

After correcting the integral term ANGINT, ANGINT1 or ANGINT2 as stated above or if it is judged that correction is not necessary in step 240, then the ECU 60 moves to a processing in step 300 shown in FIG. 4.

In step 300, the ECU 60 adds together the basic closing timing ANGBASE, the proportional term ANGPRO and the integral term ANGINT and sets the addition value (ANGBASE + ANGPRO + ANGINT) as a final closing timing ANGFIN. In the next step 302, the present final closing timing ANGFIN is set as the previous value ANGFIN0. Thereafter, the ECU 60 temporarily finishes a series of processing steps.

The ECU 60 generates a driving signal based on the final closing timing ANGFIN in another routine and outputs the driving signal to the regulation valves 70a and 60b alternately. For example, if the final closing timing ANGFIN is calculated at the timing CA1 and the timing CA3 shown in FIG. 3, the driving signal based on the final closing timing ANGFIN is outputted to the first regulation valve 70a. If the final closing timing ANGFIN is calculated at the timing CA2 and the timing CA4, the driving signal is outputted to the second regulation valve 70b. As a result, the rail pressure is controlled to coincide with the target fuel pressure PCRTRG.

It is noted here that in the above fuel pressure control, the integral terms ANGINT are updated individually for the fuel force feed systems and the final closing timing ANGFIN is calculated based on the values ANGINT1 and ANGINT2 of the integral term ANGINT corresponding to the respective fuel force feed systems.

Therefore, even if the force feed characteristics of one of the fuel force feed systems changes with the passage of time and the stationary deviation between the actual fuel pressure PCR and the target fuel pressure PCRTRG increases, only the integral value ANGINT1 or ANGINT2 of the integral term ANGINT corresponding to the fuel force feed system which force feed characteristics has changed, is updated.

Main factors for the change of the force feed characteristics of the fuel force feed systems with the passage of time involve, for example, the increase of sliding resistance between the

plungers

54a and 54b and the inner walls of the through holes 43a and 43b of the supply pumps 50a and 50b, respectively, the decrease of the response speed of the

regulation valves

70a and 70b and the elongation of response delay time from the input of the closing driving signal to the closure of the

regulation valves

70a and 70b.

If a fuel is supplied to the pressurizing

chambers

52a and 52b of the supply pumps 50a and 50b, respectively, in the suction stroke, the corresponding

plungers

54a and 54b are moved toward the cam faces 42c and the volumes of the pressurizing

chambers

52a and 52b increase. If sliding resistance increases between the

plungers

54a and 54b and the inner walls of the through holes 43a and 43b, respectively, the

plungers

54a and 54b become difficult to move and a force feed stroke is initiated without sucking the quantity of the fuel corresponding to the opening timings of the

regulation valves

70a and 70b. As a result, the fuel force feed quantities from the respective fuel force feed systems are decreased by the increase of the sliding resistance.

Furthermore, if the response delay time of the

regulation valves

70a and 70b elongates, excess fuel is sucked by the pressurizing

chambers

52a and 52b by the quantity corresponding to the response delay time even after a driving signal for opening valves is outputted to the

regulation valves

70a and 70b, respectively. Therefore, as the response delay time elongates, the fuel force feed quantities from the respective fuel force feed systems increase.

Symbol (a) in FIG. 15 indicates the change mode of the rail pressure if the force feed capability of the first fuel force feed system deteriorates due to the above-stated increase of sliding resistance and the actual fuel pressure PCR after completing the force feed operation of the pump 50a in a stationary state is always below the target fuel pressure PCRTRG.

In this case, if the integral term ANGINT is updated irrespectively of the fuel force feed systems, as shown in (b), the deviation ΔPCR1 between the actual fuel pressure PCR and the target fuel pressure PCRTRG after the completion of the fuel force feed operation of the first fuel force feed system decreases. Conversely, however, the fuel pressure of the second fuel force feed system is controlled such that the actual fuel pressure PCR after the completion of the fuel force feed operation of the second fuel force feed system always exceeds the target fuel pressure PCRTRG. As a result, the variation of the rail pressure at the time of fuel injection cannot be avoided, and the fuel injection quantity may possibly vary among the cylinders #1 to #4 or the deterioration of exhaust property may possibly be induced.

In this respect, according to the fuel pressure control in this embodiment, the integral term ANGINT is updated for each fuel force feed system even in the above case. Due to this, only the control input (final closing timing ANGFIN) of the first fuel force feed system which force feed capability has deteriorated is increases following the update of the integral term ANGINT. As shown in symbol (c) of FIG. 15, therefore, it is possible to control only the fuel pressure of the first fuel force feed system to coincide with the target fuel pressure PCRTRG without adversely influencing the fuel force feed operation of the second fuel force feed system and to suppress the irregular fuel injection quantity and the deterioration of the exhaust property.

Even if the fuel force feed quantity of one of the fuel force feed system is excess due to the elongation of the response delay time, the irregular fuel injection quantity and the deterioration of the exhaust property may occur as well. According to the fuel pressure control in this embodiment, by contrast, it is possible to suppress the occurrence of such deficiency even in that case, as well.

As can be understood from the above description, according to this embodiment, the difference in force feed characteristics between the fuel force feed systems can be reflected in the calculation of the respective control inputs (final closing timing ANGFIN). It is, therefore, possible to ensure suppressing the stationary deviation from remaining between the rail pressure and the target fuel pressure and to thereby enhance control characteristics at the time of converging the rail pressure into the target fuel pressure.

[Fourth Embodiment]

Next, the fourth embodiment of the present invention will be described. It is noted that the same constituent elements as those in the preceding embodiments will not be described herein.

In fuel pressure control in this embodiment, the proportional gain K_P is switched for the fuel force feed operation of each fuel force feed system to thereby calculate the proportional term ANGPRO in addition to the update of an integral term ANGINT for each fuel force feed system.

Now, procedures for calculating the proportional term ANGPRO will be described with reference to FIG. 16. FIG. 16 is a flow chart showing the processing content of step 450 of FIG. 13 in detail.

After step 140 shown in FIG. 13 in which the deviation ΔPCR is calculated, the ECU 60 moves to the processing in step 152 in FIG. 16. In step 152, it is judged whether the determination counter value CPCYLND is 0" or 2", i.e., whether or not a final closing timing ANGFIN to be calculated this time is intended to operate the second fuel force feed system.

If it is judged that the determination counter value CPCYLND is 0" or 2", the proportional gain K_P is set at a value K _P2 corresponding to the second fuel force feed system in step 154. On the other hand, if it is judged that determination counter value CPCYLND is 1" or 3", the proportional gain k_P is set at a value K _P1 corresponding to the first fuel force feed system in step 155.

The values K _P1 and K _P2 are preset ones based on the force feed characteristics of the respective fuel force feed systems experimentally or the like.

The force feed characteristics mentioned herein involve, for example, the fuel force feed quantities of the supply pumps 50a and 50b with respect to a predetermined final closing timing ANGFIN, the total volume of portions from the supply pumps 50a and 50b to the common rail 20 among the common rail 20 and the fuel passages and channel resistance of the respective fuel passages. If response characteristics of the fuel force feed operations of the fuel force feed systems are low such as, for example, in a case where the fuel force feed quantities are relatively low or the channel resistance is relatively high, then the values K _P1 and K _P2 of the proportional gain K_P are set relatively high. If the response characteristics of the fuel force feed systems differ from each other, therefore, the values K _P1 and K _P2 are set different from each other according to the difference in response characteristics.

After setting the proportional gain K_P as stated above, the ECU 60 calculates the proportional term ANGPRO based on the above formula (9) in step 456. Then, by executing processing in step 160 and the following shown in FIG. 2, the final closing timing ANGFIN is calculated based on the proportional term ANGPRO.

According to the fuel pressure control in this embodiment described above, even if there is a difference in response characteristics on fuel force feed operation between the fuel force feed systems, the increase quantity of the rail pressure corresponding to the predetermined control input (final closing timing ANGFIN) is almost equal. This allows the rail pressure to follow up the transitional change of the target fuel pressure at appropriate follow-up speed and control characteristics at the time of converging the rail pressure into the target fuel pressure to enhance.

Further, if the above-stated follow-up speed differs between the fuel force feed systems, there is a possibility that the pulsation of the rail pressure is induced by the variation of the follow-up speed. If such pulsation occurs to the rail pressure, irregular fuel injection quantities and the deterioration of the exhaust property may occur not only in the above-described stationary state but also in the transition state.

In this respect, according to the present invention, the rail pressure is be allowed to follow up the change of the target fuel pressure at almost constant follow-up speed and to, therefore, suppress the irregular fuel injection quantities and the deterioration of the exhaust property as described above.

In the third and fourth embodiments, the abnormality of the fuel force feed systems can be detected based on the magnitudes of the values ANGINT1 ad ANGINT2 of the integral terms ANGINT updated for the supply pumps 50a and 50b, respectively.

If abnormality is to be detected, after processing in steps 514 and 524 in FIG. 14, processing shown in the flow chart of FIG. 17 are executed prior to those in step 540 and the following.

Namely, in step 530, the absolute value |ANGINT1 - ANGINT2| of the deviation between the values ANGINT1 and ANGINT2 of the integral term ANGINT is set as an abnormality detection deviation ΔK. This abnormality detection deviation ΔK is used to judge to which fuel force feed system abnormality occurs. If the fuel force feed operation of one of the fuel force feed systems is not normally conducted, only one of the values ANGINT1 and ANGINT2 of the integral term ANGINT is increased and the abnormality detection deviation ΔK is, therefore, increased.

Next, in step 532, it is judged whether or not the abnormality detection deviation ΔK is larger than a predetermined determination value ΔK1. If the deviation ΔK is larger than the determination value ΔK1, it is judged that abnormality occurs to one of the fuel force feed systems and an abnormality flag XPUMP indicating the abnormality is set ON" in step 534.

The types of the abnormality of the fuel force feed systems detected herein involve, for example, the

stuck plungers

54a and 54b resulting from the excessive increase of the sliding resistance, malfunctions of the

regulation valves

70a and 70b and fuel leakage from the fuel passages.

After the processing in step 534 is executed or if it is judged in step 532 that the abnormality determination deviation ΔK is not more than the determination value ΔK1 and that the force feed operations of the respective fuel force feed systems are executed normally, the processing in step 540 and the following shown in FIG. 14 are executed.

If, for example, force feed capability is lowered due to the occurrence of abnormality to one of the fuel force feed systems, the fuel force feed quantity of the

other supply pump

50a or 50b is increased by feedback control so as to cancel the deviation between the rail pressure and the target fuel pressure deriving from the lowering of the force feed capability. Consequently, with a structure in which abnormality is detected based on the magnitude of the integral term ANGINT updated in common irrespectively of the fuel force feed systems, the variation width of the integral term ANGINT becomes smaller, with the result that abnormality cannot be detected or such detection is delayed.

In this respect, with the above-stated structure of the present invention, even if abnormality occurs to one of the fuel force feed systems, it is possible to detect the abnormality surely and early.

Furthermore, as yet another mode for detecting the abnormality of the respective fuel force feed systems, the absolute values |ANGINT1| and |ANGINT2| of the values ANGINT1 and ANGINT2 of the integral term ANGINT may be compared with a predetermined determination value, respectively, and the abnormality of the respective fuel force feed systems may be detected individually based on the fact that the absolute values |ANGINT1| and |ANGINT2| exceed the determination value, i.e., the values ANGINT1 and ANGINT2 of the integral term ANGINT are out of a predetermined range centered around 0".

According to the above-stated abnormality detection procedures, it is possible to identify the fuel force feed system to which abnormality occurs and to execute more accurate measures against the abnormality.

Moreover, the basic closing timing ANGBASE may be set individually for the fuel force feed systems, respectively in addition to the integral term ANGINT or the proportional term ANGPRO.

It is noted that only the proportional term ANGPRO can be set for the fuel force feed systems, respectively and that the integral term ANGINT can be set for both the fuel force feed systems in common.

Further, as a feedback control term for calculating the final closing timing ANGFIN, a differential term obtained by multiplying the differential value of the deviation ΔPCR by a predetermined gain can be used in addition to the above-stated proportional term ANGPRO and the integral term ANGINT. In that case, it is also possible to set the differential term for fuel force feed systems, respectively.

The above-stated embodiments have been described, while taking the fuel supply apparatus provided with two fuel force feed systems as an example. The present invention can be also applied to a fuel supply apparatus provided with three or more fuel force feed systems.

The above-stated embodiments have been described, taking, as an example, a diesel engine as an internal combustion engine to which the fuel pressure control device according to the present invention is applied. It is also possible to apply the fuel pressure control device according to the present invention to, for example, a direct injection gasoline engine which directly injects a fuel into a combustion chamber.

Claims

A fuel pressure control device for a high pressure fuel injection system characterized by comprising:

control means for setting a control input of a fuel pump (30) forcedly feeding a high pressure fuel to an accumulator piping (20), to which a fuel injection valve (12) is connected, based on a predetermined control gain (KFF, KFB) and thereby controlling fuel pressure (PCR) within the accumulator piping (20) so as to follow up a change of target fuel pressure (PCRTRG);

variable element detection means for detecting a variable element making a change speed of the fuel pressure within the accumulator piping (20) variable with respect to the control input at a predetermined operation speed of the fuel pump (30); and

control gain setting means (60) for variably setting the control gain (KFF, KFB) based on the variable element so as to cancel a variation of the change speed of the fuel pressure caused by the variable element in advance.
The fuel pressure control device according to claim 1, wherein the variable element is a fuel state of the high pressure fuel.
The fuel pressure control device according to claim 2, wherein

the fuel state is a bulk modulus (E) of the high pressure fuel; and

the control gain setting means (60) sets the control gain (KFF, KFB) smaller as the bulk modulus (EFB, EFF) is larger.
The fuel pressure control device according to claim 3, wherein the variable element detection means detects the bulk modulus (EFB, EFF) based on at least one of pressure (PCR) and temperature (THF) of the high pressure fuel.
The fuel pressure control device according to claim 1, wherein the variable element is a volume (VCR) of the accumulator piping.
The fuel pressure control device according to claim 1, wherein the variable element is operating characteristic of the fuel pump (30).
The fuel pressure control device according to any of claims 1 to 6, wherein the control means (60) feedback-controls the fuel pressure based on the control gain (KFF, KFB).
The fuel pressure control device according to any of claims 1 to 6, wherein the control means (60) feed-forward-controls the fuel pressure based on the control gain (KFF, KFB).
The fuel pressure control device according to one of claims 1 to 8, characterized by further comprising:

abnormality detection means (60) for detecting abnormality of a high pressure fuel injection system; and

control gain changing means (60) for forcedly changing the control gain (KFF, KFB) from a value set by the control gain setting means (60) to a value corresponding to the abnormality when the abnormality is detected.
The fuel pressure control device according to claim 9, wherein the control changing means (60) changes the control gain (KFF, KFB) to a value lower than a value set by the control gain setting means (60) when the abnormality is detected.
The fuel pressure control device according to claim 1, characterized by further comprising:

a plurality of fuel force feed systems for forcedly feeding a fuel to the fuel injection valve (12); and

control input calculation means for setting control information required according to force feed characteristics of the fuel force feed systems individually for the fuel force feed systems and calculating control inputs based on the set control information when calculating the control inputs of the fuel force feed systems to control fuel pressure of each of the plurality of fuel pressure force feed systems to coincide with target pressure (PCRTRG).
The fuel pressure control device according to claim 11, wherein the control input calculation means (60) calculates the control inputs based on feedback control.
The fuel pressure control device according to claim 12, wherein the control input calculation means (60) sets a value (ANGPRO), obtained by multiplying a deviation between actual fuel pressure (PCR) within each of the fuel force feed systems and the target pressure (PCRTRG) by a proportional gain (K_P) set for each of the fuel force feed systems, as the control information.
The fuel pressure control device according to claim 12 or 13, wherein the control input calculation means (60) sets a value (ANGINT), obtained by integrating a deviation between actual fuel pressures (PCR) within each of the fuel force feed systems and the target fuel pressure (PCRTRG) for each of the fuel force feed systems, as the control information.
A Method for controlling fuel pressure in a high pressure fuel injection system characterized by the steps:

setting a control input of a fuel pump (30) forcedly feeding a high pressure fuel to an accumulator piping (20), to which a fuel injection valve (12) is connected, based on a predetermined control gain (KFF, KFB) and thereby controlling fuel pressure (PCR) within the accumulator piping (20) so as to follow up a change of target fuel pressure (PCRTRG);

detecting a variable element making a change speed of the fuel pressure within the accumulator piping (20) variable with respect to the control input at a predetermined operation speed of the fuel pump (30) ; and

variably setting the control gain (KFF, KFB) based on the variable element so as to cancel a variation of the change speed of the fuel pressure caused by the variable element in advance.
The method according to claim 15, wherein the variable element is a fuel state of the high pressure fuel.
The method according to claim 16, wherein

the fuel state is a bulk modulus (E) of the high pressure fuel; and the further step:

setting the control gain (KFF, KFB) smaller as the bulk modulus (EFB, EFF) is larger.
The method according to claim 17, wherein the bulk modulus (EFB, EFF) is detected based on at least one of pressure (PCR) and temperature (THF) of the high pressure fuel.
The method according to claim 15, wherein the variable element is a volume (VCR) of the accumulator piping.
The method according to claim 15, wherein the variable element is operating characteristic of the fuel pump (30).
The method according to any of claims 15 to 20, wherein the fuel pressure is feedback-controlled based on the control gain (KFF, KFB).
The method according to any of claims 15 to 20, wherein the fuel pressure is feed-forward-controlled based on the control gain (KFF, KFB).
The method according to one of claims 15 to 22, characterized by the further steps:

detecting abnormality of a high pressure fuel injection system; and

forcedly changing the control gain (KFF, KFB) from a value set to a value corresponding to the abnormality when the abnormality is detected.
The method according to claim 23, with the further step of changing the control gain (KFF, KFB) to a value lower than a value set when the abnormality is detected.
The method according to claim 15, characterized by the further steps:

forcedly feeding a fuel to the fuel injection valve (12) by a plurality of fuel force feed systems; and

setting control information required according to force feed characteristics of the fuel force feed systems individually for the fuel force feed systems and calculating control inputs based on the set control information when calculating the control inputs of the fuel force feed systems to control fuel pressure of each of the plurality of fuel pressure force feed systems to coincide with target pressure (PCRTRG).
The method according to claim 25, wherein the control inputs are calculated based on feedback control.
The method according to claim 26, wherein a value (ANGPRO) is set, obtained by multiplying a deviation between actual fuel pressure (PCR) within each of the fuel force feed systems and the target pressure (PCRTRG) by a proportional gain (K_P) set for each of the fuel force feed systems, as the control information.
The method according to claim 26 or 27, wherein a value (ANGINT) is set, obtained by integrating a deviation between actual fuel pressures (PCR) within each of the fuel force feed systems and the target fuel pressure (PCRTRG) for each of the fuel force feed systems, as the control information.