CN103970099A - Flying robot multi-sensor scheduling system and method for overhead power transmission line patrolling - Google Patents

Abstract

The invention discloses a flying robot multi-sensor scheduling system and method for overhead power transmission line patrolling, and relates to the technical field of detection and control of power system power transmission lines. The scheduling system comprises an action planning layer, a time-sharing scheduling layer and a sensor control layer, wherein the action planning layer, the time-sharing scheduling layer and the sensor control layer are sequentially connected. The scheduling method includes the steps that an original value of a set A* of sensors to be selected is set to be a null set; a set G is set to meet the equation G=S-A*, if the set G is not the null set, a sensor S' is selected from the set G so that the selected sensor S' meets the condition in the specification, operation cost c (A') of a set A' is calculated, and when c(A') is smaller than or equal to cL, the set A is set to be equal to A' and the set G is set to meet the equation that G=G-S'; when the set G is a null set, the sensors in the set A of the sensors to be selected are scheduled sensors. The system and method solve the problem that in an existing patrolling system, the sensors cannot be selected automatically according to different patrolled objects, and patrolling efficiency is improved.

Description

Overhead transmission line is patrolled and examined flying robot's multisensor dispatching system and method

Technical field

The invention belongs to power system transmission line and detect and control technology field, relate in particular to a kind of overhead transmission line and patrol and examine flying robot's multisensor dispatching system and method.

Background technology

Patrolling and examining of transmission line of electricity is an element task that effectively guarantees transmission line of electricity and device security thereof.There is potential safety hazard and patrol and examine the impact that result is subject to too much human factor in traditional manual inspection.Patrol and examine flying robot carry multiple sensors flight along the line to each parts of circuit patrol and examine efficient height, safety coefficient is high and the advantage such as applicable field work, has made up to a great extent the deficiency of manual inspection; Flying robot's operation and maintenance cost is low, has also made up the high shortcoming of helicopter routing inspection use cost.

Manned aircraft line walking or unmanned plane line walking in the past all adopt multisensor parallel detection mode, to different task, different target, all use all the sensors jointly to detect.Although this can guarantee maximum detectability, the unnecessary wasting of resources and the redundancy of data have also been caused.

Select optimum sensor group to patrol and examine and can save computing time, reduce physics loss, avoid unnecessary sensor action, improve and patrol and examine efficiency, reduce data redundancy, guarantee the completeness of data acquisition, improve system reliability and robustness.

Because sensor type increases, it is wide that range of application becomes, and it is more complicated that function also becomes, and simply relies on manpower cannot complete sensor management work.

Summary of the invention

The object of the invention is to, provide a kind of overhead transmission line to patrol and examine flying robot's multisensor dispatching system and method, for solving existing cruising inspection system, cannot independently select according to the difference of patrolling and examining object the problem of various kinds of sensors, make cruising inspection system detect the sensor combinations of target selection optimum and appropriately to determine each sensor control parameter each, improve and patrol and examine efficiency.

To achieve these goals, the technical scheme that the present invention proposes is that a kind of overhead transmission line is patrolled and examined flying robot's multisensor dispatching system, it is characterized in that described dispatching system comprises action planning layer, timesharing dispatch layer and sensor key-course that order is connected;

Described action planning layer patrols and examines region for determining, patrol and examine type, patrol and examine target and patrol and examine the status information of target;

Described timesharing dispatch layer for according to patrolling and examining region, patrol and examine type, patrol and examine target and patrol and examine the status information of target, solve optimal sensor set;

Described sensor key-course carries out target for the sensor of dispatching optimal sensor set and patrols and examines.

Overhead transmission line is patrolled and examined flying robot's multisensor dispatching method, it is characterized in that described dispatching method comprises:

Step 1: make set of sensors A to be selected ^*initial value be empty set;

Step 2: order set G=S-A ^*, the set of S for being formed by all the sensors; Whether judgement set G is empty set, if set G is empty set, performs step 7; Otherwise, execution step 3;

Step 3: choose a sensor S ' in set G, the sensor S ' choosing is met $\arg \underset{A^{\′}\∈A}{\max }\left\{f\left(A^{\′}\right)\right\};$

Wherein, set A '=A ^*∪ S ';

A is the set that the various combinations by all the sensors form;

f(A′)＝I(θ；A′)＝H(θ)-H(θ|A′)；

θ is the status information of patrolling and examining target;

H (θ) for patrol and examine target status information entropy and

θ _ifor patrolling and examining i status information of target;

P (θ _i) be to patrol and examine the probability of target when i status information;

N is the state number of patrolling and examining target;

$H\left(\mathrm{\θ}\right|A^{\′})=-\underset{\mathrm{\θ}}{\mathrm{\Σ}}\underset{S_1}{\mathrm{\Σ}}\underset{S_2}{\mathrm{\Σ}}...\underset{S_n}{\mathrm{\Σ}}\left\{P(\mathrm{\θ},S_1,S_2,...,S_n){\log }_{2}P\left(\mathrm{\θ}\right|S_1,S_2,...,S_n)\right\};$

S _ifor set A ' in i sensor;

P (θ, S ₁, S ₂..., S _n) be (θ, S ₁, S ₂..., S _n) between joint probability distribution;

P (θ | S ₁, S ₂..., S _n) be conditional probability;

N be set A ' in the number of sensor;

The operating cost c of step 4: set of computations A ' (A '), if c (A ')≤c _l, perform step 5; Otherwise, execution step 6;

Wherein, c _lfor setting threshold;

Step 5: make set A ^*=A ';

Step 6: order set G=G-S ', and return to step 2;

Step 7: set of sensors A to be selected ^*in sensor for scheduling sensor.

Described operating cost comprises energy consumption cost, carrying cost, assesses the cost and communications cost.

Described set A ' the computing formula of operating cost c (A ') be:

$c\left(A^{\′}\right)=w_1\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{e}_i+w_2\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{i}_i^s+w_{3}n+w_4\frac{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{i}_i^s}{\mathrm{BW}};$

Wherein, e _ifor set A ' in the energy consumption cost of i sensor;

W ₁for the energy consumption cost of the sensor weight in operating cost;

for set A ' in i sensor take the shared carrying cost of single image;

W ₂for sensor is taken the weight of the shared carrying cost of single image in operating cost;

W ₃the weight assessing the cost in operating cost for sensor;

BW is the network bandwidth;

W ₄for the communications cost of the sensor weight in operating cost;

N be set A ' in the number of sensor.

The invention solves sensor in existing cruising inspection system and cannot according to the difference of patrolling and examining object, independently select the problem of various kinds of sensors, improved and patrolled and examined efficiency, improved reliability and the robustness of system.

Accompanying drawing explanation

Fig. 1 is that overhead transmission line provided by the invention is patrolled and examined flying robot's multisensor dispatching system structural drawing;

Fig. 2 is that overhead transmission line provided by the invention is patrolled and examined flying robot's multisensor dispatching method process flow diagram.

Embodiment

Below in conjunction with accompanying drawing, preferred embodiment is elaborated.Should be emphasized that, following explanation is only exemplary, rather than in order to limit the scope of the invention and to apply.

Fig. 1 is that overhead transmission line provided by the invention is patrolled and examined flying robot's multisensor dispatching system structural drawing.As shown in Figure 1, dispatching system provided by the invention comprises action planning layer, timesharing dispatch layer and the sensor key-course that order is connected, and every one deck completes relatively independent function, a plurality of functional modules or assembly, consists of.

Action planning layer is formulated total patrol plan by patrol officer, determines and patrols and examines region, patrols and examines type and patrol and examine target, simultaneously for Sensor scheduling provides background information, for patrolling and examining target, provides state hypothesis, i.e. status information.Wherein, patrol and examine target and include but not limited to shaft tower, line of electric force and insulator etc., and the status information of patrolling and examining target includes but not limited to normal condition and abnormality.

Timesharing dispatch layer is the core layer of whole dispatching system, its main modular is Sensor scheduling module, this module is carried out sensor optimization coordinated scheduling according to optimal scheduling algorithm, its objective is and in limited cost budgeting, solves optimal sensor set, for system provides maximum information increment.

Sensor key-course, in the bottom, its role is to the control parameter of each sensor in optimal sensor set suitably to adjust, and realizes sensor optimal objective and patrols and examines, and improves to the full extent the performance of system.

The present invention also provides a kind of overhead transmission line to patrol and examine flying robot's multisensor dispatching method, and the key step of the method realizes at timesharing dispatch layer.Fig. 2 is that overhead transmission line provided by the invention is patrolled and examined flying robot's multisensor dispatching method process flow diagram.As shown in Figure 2, dispatching method provided by the invention comprises:

Step 1: make set of sensors A to be selected ^*initial value be empty set.Set of sensors A to be selected ^*for storing the final sensor that needs scheduling.

Step 2: order set G=S-A ^*.

If S={S ₁, S ₂..., S _nbe the set being formed by all the sensors, S _ifor all the sensors and the i=1 of dispatching system control, 2 ..., n, n is the quantity of all the sensors of dispatching system control.

Whether judgement set G is empty set, if set G is empty set, performs step 7; Otherwise, execution step 3.

Step 3: make A={A ₁, A ₂... A _mit is the set that the various combinations by all the sensors form.Such as, if S={S ₁, S ₂, S ₃, can there is A ₁={ S ₁, A ₂={ S ₂, A ₃={ S ₃, A ₄={ S ₁, S ₂, A ₅={ S ₁, S ₃, A ₆={ S ₂, S ₃, A ₇={ S ₁, S ₂, S ₃.And A={A ₁, A ₂... A ₇be exactly by sensor S ₁, S ₂and S ₃the set that forms of various combinations.

In set G, choose a sensor S ', the sensor S ' choosing met:

$\arg \underset{A^{\′}\∈A}{\max }\left\{f\left(A^{\′}\right)\right\}---\left(1\right)$

In formula (1), set A '=A ^*∪ S ', objective function f (A ') is system information increment, can use mutual information I (θ; A ') weigh, the computing formula of f (A ') is:

f(A′)＝I(θ；A′)＝H(θ)-H(θ|A′) (2)

In formula (2), θ is the status information of patrolling and examining target, and H (θ) is for to patrol and examine the status information entropy of target and to meet formula:

$H\left(\text{\θ}\right)=-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}p\left({\mathrm{\θ}}_i\right){\log }_{2}p\left({\mathrm{\θ}}_i\right)---\left(3\right)$

In formula (3), θ _ifor patrolling and examining i status information of target, p (θ _i) be to patrol and examine the probability of target when i status information, N is the state number of patrolling and examining target.

And H (θ | A ') is conditional entropy, represent to obtain set A ' after, patrol and examine the average information of the status information of target, i.e. average uncertainty.The computing formula of H (θ | A ') is as follows:

$H\left(\mathrm{\θ}\right|A^{\′})=-\underset{\mathrm{\θ}}{\mathrm{\Σ}}\underset{S_1}{\mathrm{\Σ}}\underset{S_2}{\mathrm{\Σ}}...\underset{S_n}{\mathrm{\Σ}}\left\{P(\mathrm{\θ},S_1,S_2,...,S_n){\log }_{2}P\left(\mathrm{\θ}\right|S_1,S_2,...,S_n)\right\}---\left(4\right)$

In formula (4), S _ifor set A ' in i sensor, P (θ, S ₁, S ₂..., S _n) be (θ, S ₁, S ₂..., S _n) between joint probability distribution, P (θ | S ₁, S ₂..., S _n) be conditional probability, n be set A ' in the number of sensor.

The operating cost c of step 4: set of computations A ' (A ').

In the present invention, the effect of the Sensor scheduling module of action planning layer is in certain operating cost, to select optimal sensor combination.Wherein operating sensor cost comprises energy loss, computing time, storage space and communication bandwidth.For each sensor, there are energy consumption cost, carrying cost, assess the cost and these several costs of enabling of communications cost.Energy consumption equals the power consumption of sensor device, and carrying cost equals the size of individual picture of sensor shooting, assesses the cost and is directly proportional to number of sensors, and communications cost is transmission required minimum time of single image (being the ratio of image size and bandwidth).Above-mentioned four kinds of costs are added and can obtain the total operating cost of sensor group by power.If above-mentioned four kinds of shared weights of cost are respectively w ₁～w ₄, set A ' the computing formula of operating cost be:

$c\left(A^{\′}\right)=w_1\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{e}_i+w_2\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{i}_i^s+w_{3}n+w_4\frac{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{i}_i^s}{\mathrm{BW}}---\left(5\right)$

In formula (5), e _ifor set A ' in the energy consumption cost of i sensor, for set A ' in i sensor carrying cost of taking single image.Make set A ' in assessing the cost of single-sensor be 1, each sensor assesses the cost identical, set A ' in the sum that assesses the cost of all the sensors be n.Meanwhile, n be also set A ' in the number of sensor.BW is the network bandwidth.

Setting operating cost threshold value is c _l, judgement c (A ')≤c _lwhether set up, if c (A ')≤c _l, perform step 5; Otherwise, execution step 6.

Step 5: make set A ^*=A '.Because the sensor S' choosing meets be that system information increment increases, and meet c (A ')≤c _l, operating cost is in setting threshold, and the sensor S' therefore choosing meets the condition that becomes optimal sensor combination, so make set A ^*=A ', even sensor S' becomes set of sensors A to be selected ^*in element.

Step 6: order set G=G-S '.Because sensor S' is once processed, therefore no matter whether sensor S' meets is put into set of sensors A to be selected ^*in condition, all it should be deleted G from set.Then, return to step 2, process other elements in set G.

Step 7: set of sensors A to be selected ^*in sensor for scheduling sensor.

Due to set of sensors A to be selected ^*in, be all to meet system information increment to increase, and the element of operating cost in setting threshold, therefore set of sensors A to be selected ^*in element be optimum scheduling sensor.Sensor key-course is dispatched set of sensors A to be selected ^*in sensor, comprise unlatching, focusing, gain-adjusted etc., realize Sensor scheduling, thereby improve, patrol and examine efficiency.

The invention solves sensor in existing cruising inspection system and cannot independently select according to the difference of patrolling and examining object the problem of various kinds of sensors, reduced physics loss, avoided unnecessary sensor action, improved and patrolled and examined efficiency, reduced data redundancy, guarantee the completeness of data acquisition, improved system reliability and robustness.

The above; be only the present invention's embodiment preferably, but protection scope of the present invention is not limited to this, is anyly familiar with in technical scope that those skilled in the art disclose in the present invention; the variation that can expect easily or replacement, within all should being encompassed in protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.

Claims (4)

1. overhead transmission line is patrolled and examined flying robot's multisensor dispatching system, it is characterized in that described dispatching system comprises action planning layer, timesharing dispatch layer and sensor key-course that order is connected;

Described action planning layer patrols and examines region for determining, patrol and examine type, patrol and examine target and patrol and examine the status information of target;

Described timesharing dispatch layer for according to patrolling and examining region, patrol and examine type, patrol and examine target and patrol and examine the status information of target, solve optimal sensor set;

Described sensor key-course carries out target for the sensor of dispatching optimal sensor set and patrols and examines.

2. overhead transmission line is patrolled and examined flying robot's multisensor dispatching method, it is characterized in that described dispatching method comprises:

Step 1: make set of sensors A to be selected ^*initial value be empty set;

Step 2: order set G=S-A ^*, the set of S for being formed by all the sensors;

Whether judgement set G is empty set, if set G is empty set, performs step 7; Otherwise, execution step 3;

Step 3: choose a sensor S ' in set G, the sensor S ' choosing is met $\arg \underset{A^{\′}\∈A}{\max }\left\{f\left(A^{\′}\right)\right\};$

Wherein, set A '=A ^*∪ S ';

A is the set that the various combinations by all the sensors form;

f(A′)＝I(θ；A′)＝H(θ)-H(θ|A′)；

θ is the status information of patrolling and examining target;

H (θ) for patrol and examine target status information entropy and

θ _ifor patrolling and examining i status information of target;

P (θ _i) be to patrol and examine the probability of target when i status information;

N is the state number of patrolling and examining target;

$H\left(\mathrm{\θ}\right|A^{\′})=-\underset{\mathrm{\θ}}{\mathrm{\Σ}}\underset{S_1}{\mathrm{\Σ}}\underset{S_2}{\mathrm{\Σ}}...\underset{S_n}{\mathrm{\Σ}}\left\{P(\mathrm{\θ},S_1,S_2,...,S_n){\log }_{2}P\left(\mathrm{\θ}\right|S_1,S_2,...,S_n)\right\};$

S _ifor set A ' in i sensor;

P (θ, S ₁, S ₂..., S _n) be (θ, S ₁, S ₂..., S _n) between joint probability distribution;

P (θ | S ₁, S ₂..., S _n) be conditional probability;

N be set A ' in the number of sensor;

The operating cost c of step 4: set of computations A ' (A '), if c (A ')≤c _l, perform step 5; Otherwise, execution step 6;

Wherein, c _lfor setting threshold;

Step 5: make set A ^*=A ';

Step 6: order set G=G-S ', and return to step 2;

Step 7: set of sensors A to be selected ^*in sensor for scheduling sensor.

3. dispatching method according to claim 2, is characterized in that described operating cost comprises energy consumption cost, carrying cost, assesses the cost and communications cost.

4. dispatching method according to claim 3, it is characterized in that described set A ' the computing formula of operating cost c (A ') be:

$c\left(A^{\′}\right)=w_1\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{e}_i+w_2\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{i}_i^s+w_{3}n+w_4\frac{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{i}_i^s}{\mathrm{BW}};$

Wherein, e _ifor set A ' in the energy consumption cost of i sensor;

W ₁for the energy consumption cost of the sensor weight in operating cost;

for set A ' in i sensor take the shared carrying cost of single image;

W ₂for sensor is taken the weight of the shared carrying cost of single image in operating cost;

W ₃the weight assessing the cost in operating cost for sensor;

BW is the network bandwidth;

W ₄for the communications cost of the sensor weight in operating cost;

N be set A ' in the number of sensor.