WO2013071321A2

WO2013071321A2 - Method for regulating power streams

Info

Publication number: WO2013071321A2
Application number: PCT/AT2012/000290
Authority: WO
Inventors: Andreas Kuhn; Birgit Kuhn
Original assignee: Andreas Kuhn; Birgit Kuhn
Priority date: 2011-11-14
Filing date: 2012-11-14
Publication date: 2013-05-23
Also published as: WO2013071321A3; AT512133A1

Abstract

The invention relates to a method for regulating power streams in a power supply system (1) having a plurality of hierarchically arranged units (91, 92, 93, 94, 95), which are connected by means of lines (8), wherein the plurality of units (91, 92, 93, 94, 95) each have at least one associated regulatory element (2), wherein the regulatory elements (2) are essentially coupled in accordance with the lines (8) of the associated units (91, 92, 93, 94, 95) and are arranged in the region of the respective associated unit (91, 92, 93, 94, 95), wherein two regulatory elements (2) coupled to one another are designed such that at least one portion of the manipulated variable (41) of the hierarchically superordinate regulatory element (2) is prescribed as a target selection (31) for the hierarchically subordinate regulatory element (2), and the regulation is effected in subsidiary fashion.

Description

Method for controlling energy flows

The invention relates to a method for controlling energy flows in a power supply network according to the preamble of claim 1.

Such power grids have in the past had some large power producers whose power has been delivered to consumers via lines. In contrast, modern energy supply networks have a heterogeneous structure both on the side of the energy producers and on the side of the consumers. In particular, there are a large number of small energy producers, for example wind power plants, solar plants, district heating plants or plants for the use of other alternative and regenerative energy sources. Furthermore, the number and complexity of devices in households and industry is increasing.

The regulation and coordination of this multitude of units is a great challenge, for a variety of different solutions

were proposed. However, it has been shown in practice that these solutions can often insufficiently integrate individual units into the overall system in a system with a large number of heterogeneous units.

Object of the present invention is to provide a method of the type mentioned, in which the known disadvantages are avoided, which can be easily implemented and expanded, allows a good regulation with a low resource requirements and can be adapted quickly and automatically to different conditions ,

This is achieved by the features of claim 1 according to the invention.

This has the advantage that an efficient regulation of

Power supply network can be provided, which is decentralized and can respond to the needs of each unit. The required data transfer can be kept low because the scheme is delegated as far as possible to subordinate hierarchy levels. The control can effectively avoid peak loads and ensure good coordination of different units. Is the energy supply network an electrical one?

Energy supply network, so it represents an intelligent power grid, for which the keyword "Smart Grid" is common.

The invention also relates to a power supply control system for controlling energy flows in a power supply network with which the method described above can be implemented particularly easily.

The invention further relates to a computer program product which can be loaded directly into the internal memory of a computer and / or a computer network and comprises software code sections with which the steps of the method described at the outset are carried out

Computer program product on a computer and / or one

Computer network is running.

The subclaims relate to further advantageous embodiments of the invention.

It is expressly referred to the wording of the claims, whereby the claims are hereby incorporated by reference into the description and are considered to be reproduced verbatim.

The invention will be described in more detail with reference to the accompanying drawings, in which only preferred embodiments are shown by way of example. Showing:

Fig. 1 is a schematic representation of a power supply network; and

Fig. 2 is a schematic representation of a control element for the

Power grid.

Fig. 1 shows schematically a power supply network 1. The

Power supply network 1 has a heterogeneous structure of units 91, 92, 93, 94, 95. The units 91, 92, 93, 94, 95 are energy producers,

Consumers, distributors, memory od. Like. Be. It can also be provided that a unit 91, 92, 93, 94, 95 can assume a plurality of functions, for example, as a memory or generator. Thus, an accumulator of an electric car is usually operated so that it energy from the

Energy supply network 1 refers and thus acts as a consumer. However, it can also be provided that the accumulator is operated as a memory and, if necessary, energy is returned to the energy supply network 1.

Further embodiments of the units 91, 92, 93, 94, 95 may be energy producers, such as nuclear power plants, wind turbines or solar systems, or

Distributors, for example substations, storage, for example accumulators, pumped storage power plants or hot water storage, or consumers,

For example, factory machinery, household appliances, lifts, be. The units

91, 92, 93, 94, 95 can also be used, for example,

Heat pumps or water boilers concern.

This invention is concerned with the regulation of energy flows in such power supply networks 1 in which a plurality of units 91, 92, 93, 94, 95 connected by means of lines 8 are present, the units 91,

92, 93, 94, 95 are arranged hierarchically.

The energy supply network 1 can relate to networks for the provision of different energy sources, wherein several different energy sources can be provided. In particular, units 91, 92, 93, 94, 95 for energy conversion can also be provided. For example, a

Pumped storage power plant convert electrical energy into potential energy and vice versa.

The power grid can be a gas supply network, a

Heat supply network or the like. A preferred embodiment of the power supply network 1 relates to an electrical power supply network.

Other embodiments of the power grid 1 relate to district heating, the distribution of fossil fuels, geothermal or the like.

If the hierarchical structure is considered from bottom to top, then

For example, the appliances in a household, in particular a washing machine, a refrigerator, a stove, a dishwasher, a heater, the next higher unit may be the apartment distribution unit 92, which is superior to the floor distribution unit 93, which is the parent house distribution unit 95 further. In the hierarchical structure, in the

Substantially in each level, there may be provided directly units 91, 92, 93, 94, 95 which are power generators, consumers, distributors, storages. For example, the house distribution unit 95 next to several floor distribution units 93, a further unit 94 may be directly subordinated, which may be, for example, a swimming pool heating or solar system. Such hierarchical structures exist in all areas of energy supply, such as industry or public space.

The number of hierarchy levels is not restricted and can be selected according to the respective energy supply network.

Typically, in the power grid 1, a plurality of the units 91, 92, 93, 94, 95 energy producers, such as nuclear power plants, hydroelectric power plants, wind turbines or solar systems.

In the power grid 1 energy producers and energy consumers are combined, with both energy producers and energy consumers can be present in one level.

The plurality of units 91, 92, 93, 94, 95 are each at least one

Rule element 2 assigned, wherein the control elements 2 substantially

are coupled according to the lines 8 of the associated units 91, 92, 93, 94, 95 and in the region of the respectively associated unit 91, 92, 93, 94, 95 are arranged.

Two mutually coupled control elements 2 are designed such that at least a part of the manipulated variable 41 of the hierarchically superior

Control element 2 as a target 31 'of the hierarchically subordinate

Control element 2 is specified and the regulation is subsidiary.

In particular, the plurality of control elements 2 coupled to each other can be formed in this way. A subsidiary regulation is a regulation which takes place decentrally and the regulation is delegated as far as possible to hierarchically subordinate structures.

This solves problems directly in the affected areas.

In particular, with the subsidiary regulation, the required flow of information can be kept low. The power grid control system breaks down into individual subsystems, which have lower complexity and can be solved more easily with little effort.

In the case of a subsidiary rule, the required regulatory intelligence is arranged as deeply as possible and as high as necessary, from a hierarchical point of view.

Due to this design of the control elements 2, it is not necessary that the energy supplier must grasp the energy supply network to the last realization depth. For example, he does not have to regulate the individual dishwasher, but regulates, for example, simply about his pricing model, which he can bring in an upper hierarchical level.

The control elements 2 form a power grid control system for controlling energy flows in the power grid 1 with a plurality of hierarchically arranged units 91, 92, 93, 94, 95 connected by lines 8, wherein the plurality of units 91, 92, 93, 94, 95 respectively at least one of the control elements 2 is assigned, wherein the control elements 2 substantially according to the lines 8 of the associated units 91, 92, 93, 94, 95 and coupled in the region of the respectively associated unit 91, 92, 93, 94, 95 are arranged. Two mutually coupled control elements 2 are designed such that the manipulated variable 41 of the hierarchically superior

Control element 2, the target 31 'of the hierarchically subordinate

Rule element 2 includes and the regulation is subsidiary.

An efficient regulation of the energy supply network 1 can be provided, which takes place locally and can respond to the needs of the individual units 91, 92, 93, 94, 95. The required data transfer can be kept low because the scheme as far as possible to subsequent

Hierarchy levels is delegated. Load peaks can be effectively avoided and good coordination of different units 91, 92, 93, 94, 95 be ensured.

The power grid control system reacts like an automated energy market, with market pressure regulating the complex system essentially without central intelligence.

Is it in the power grid 1 to an electrical

Due to the heterogeneous structure of the units 91, 92, 93, 94, 95 combined with a large number of different setting possibilities, the coordination and regulation of the energy supply network 1 represents a great challenge. It should be noted that different units 91, 92, 93, 94 , 95 have different dynamics and behavioral characteristics. By small solar systems, wind, hydropower, biogas plants and the like, the power grid 1 has a large number of different energy producers.

In particular, energy producers, which are dependent on the weather, have a stochastic energy supply and can only be regulated to a limited extent.

The present method and the power grid control system can be easily extended to other units 91, 92, 93, 94, 95, wherein the existing power grid control system can remain substantially unchanged. The extensions may be the addition of further hierarchically subordinate units 91, 92, 93, 94, 95 to already integrated control elements 2. But there can be more

Hierarchy levels are introduced and, for example, several regional

Power supply control systems can be merged, whereby a global power grid control system can be created.

The present method may simply be incremental to an existing one

Power supply network 1 are introduced, the continuous operation of the power supply network 1 can be largely ensured.

The control elements 2 may be substantially the same in each hierarchical level be constructed, which may be provided, largely similar

To use control elements 2 for the majority of control elements 2, which in particular the development and manufacturing costs can be kept low.

As low has been found, if at least one, in particular the majority, the control elements 2, a machine-learning model is used.

At least one, in particular the majority, of the control elements 2 comprises a machine learning model. This can ensure that the

Control element 2, the power grid 1 can map well in its area, with an analytical presentation is not required and the desired behavior of the control element 2 can be trained.

Machine learning models can be, for example, artificial neural networks, regression or classification trees, support vector machines and / or look-up tables and / or committees of machine-learning models.

The machine learning model can be adaptively adapted to the power grid 1. As a result, both the accuracy and reliability of the control can be increased, as well as a substantially automatic adaptation to changing conditions can be achieved.

The control elements 2 can be designed to be both adaptive and adaptive in terms of consumer behavior as well as with regard to the provider side. In this case, a forecast of the price development can take place, for example, that in fair weather electricity is offered cheaper because the solar cells provide more energy.

By using adaptive control elements 2 it can be ensured that an adaptation to different goals and circumstances, for example a changed consumer behavior, a new generation of devices or environmental influences, can be done quickly and easily and despite this

Changes a good regulatory outcome can be achieved.

The adaptation can be used for new situations on the part of energy producers,

Being a consumer, distributor or storage. In particular, an adaptation to Consumption patterns and / or producer pattern done.

It can be provided that the control element 2 is trained, which control possibilities with individual units 91, 92, 93, 94, 95 are possible. For example, a control element 2, to which a solar system and a

Heat pump of a hot water tank are connected, use the solar energy to heat the hot water tank. But it can also be provided that the solar energy is fed into the power grid 1 and the hot water tank at a later date, for example, at night when the power of the solar system is significantly reduced, is heated with energy from the power grid 1. It can also be provided that the hot water tank is overheated in a tolerance range and at a later time, for example, in the evening when the power of the solar system is significantly reduced, by means of the heat pump energy in the

Power supply network 1 is fed. In this case, the most uniform possible utilization of the power supply network 1 can be ensured, in particular an overload of the power supply network 1 can be avoided.

It can be provided that the units 91, 92, 93, 94, 95 of the associated control unit 2 transmit information about the possible areas of application.

2, an embodiment of a control element 2 is shown schematically. The control element 2 can be used in all hierarchy levels and can essentially always be constructed the same.

In the method, the control element 2, a target 31 can be specified. Furthermore, 2 energy supply network-related input parameters 33 are obtained from the control element, after which 2 model parameters of the current power supply are determined by the control element, and then under

Taking into account the model parameters and the target 31 manipulated variables 41 are determined.

For a unit 91, 92, 93, 94 95, in particular a network operator, can as a manipulated variable 41 and / or as a target 31 to hierarchically subordinate Control elements 2 a price specification can be selected. The price specification can be used as a parameter for the target 31 at lower hierarchy levels.

It may also be provided that the pricing is in the form of a dynamic pricing model, for example, the prices based on the

Consumption forecasts can be given dynamically to the

Influence consumption patterns.

In addition to the manipulated variable 41, a demand message 43 is also preferably determined, which is transmitted to the hierarchically higher-order control element 2. This demand message 43 may include the urgency and options of the demand. For example, the required duration and completion time of a unit 91, 92, 93, 94, 95 may be included in the requirement message 43, resulting in possible start times.

In this case, two control elements 2 coupled to one another can be designed such that a requirement message 43 of the hierarchically subordinate one

Control element 2 is transmitted to the hierarchically superior control element 2.

Price or general costs can also be parameters of the target 31 and / or the requirement message 43, which can be used for the communication between the hierarchy levels purposefully. Thus, the energy flows in the power grid 1 can level off by itself.

The power grid related input parameters 33 may include a

Requirement 34, which results directly from the requirement message 43 of the hierarchically subordinate rule element 2. The requirement messages 43 are all hierarchically directly subordinate

Control elements 2 combined to the requirement 34. Furthermore you can

Sensor data of the situation of the power supply network 1 may be included. The power supply related input parameters 33 may also include scheduled failures due to maintenance or the like.

In this case, the requirement messages 43 of the hierarchically subordinate

Control elements 2, the requirement 34 as part of power supply-related input parameters 33 formed. The request 34 may be directly from the requirement messages 43 of the hierarchical

Subordinate rule elements 2 result or in the hierarchical

Parent control element 2, which the requirements messages 43 are passed, are formed.

The model parameters of a model device 51 of the

Control element 2, and the manipulated variables 41 are determined by a detection device 52 of the control element 2.

The control element 2, a model device 51 and a

Detection device 52 include, wherein the pattern device 51 and the detection device 52 are formed as a separate interfaced by means of modules modules.

With the model device 51, the situation of the power supply network 1 can be identified and / or predicted. With the determination device 52, the best possible action and / or control parameters can be determined.

By providing the model device 51 and the determination device 52, the control element 2 can be subdivided into independent subsystems, which enable a simpler design of the control element 2. In this case, the model device 51 and the determination device 52 in

Essentially independently adapted and trained.

By means of suitable feedback, the results of the determination device 52 can be taken into account by the model device 51.

In at least one, in particular the majority, of the control elements 2, both for the model device 51 a machine learning model and for the determination device 52 a machine learning model can be used. In this case, the machine learning model of the model device 51 can be designed differently from the machine learning model of the determination device 52.

The model device 51 may comprise a machine-learning model for the identification and / or prognosis of the situation and the determination device 52 may a machine-learning model for the determination of the manipulated variables 41 include.

The model device 51 can serve to identify and forecast the situation, for example an increasing need, an immediate danger of a situation

Overload peak or the offer of cheap electricity in the next four hours. The determination device 52 can serve to regulate the actual action as best as possible for the local circumstances and possibilities.

In particular, in the determination device 52 may be that several different machine-learning models are provided, depending on what state the system is currently or which

Strategy is pursued, for example, as quickly as possible heating and / or cooling a room, saving energy without freezing people in the room, avoiding temperature fluctuations.

Furthermore, it can be provided that a control strategy specification is determined by means of a judging means 53 and the manipulated variables 41 are determined by the control unit

Determination device 52 are determined taking into account the control strategy specification.

The control element 2 may further comprise a judging means 53, wherein the judging means 53 is formed as a stand-alone module coupled by means of interfaces.

By the judging means 53 can easily and quickly the

Regulierungsstrategievorgabe to the respective situation of the control element 2 associated unit 91, 92, 93, 94, 95 are adjusted. In particular, it can be avoided that due to a local problem area, a global strategy change is required, which also has an impact on other local areas and can possibly lead to problems there.

The manipulated variables can be targets 31 'for hierarchically subordinate

Control elements 2 and / or operating parameters 42 for the control element 2 associated unit 91, 92, 93, 94, 95 include.

From the control element 2 can be independent of power supply network Input parameters 32 are related, which are taken into account in the determination of the manipulated variables 41. The inclusion of

power supply independent input parameter 32 can significantly improve the quality of the control, whereby for the power grid 1 indirect factors can be easily taken into account.

From the model device 51 further model parameters of the future energy supply can be determined and the further model parameters can be used for the determination of the manipulated variables 41.

In particular, energy supply network-independent input parameters 32 can be taken into account when determining the model parameters and / or the further model parameters.

The power supply independent input parameters 32 may

For example, weather data and / or calendar data include. In this way, the change in demand due to a holiday can be easily and quickly taken into account. Furthermore, based on the weather data, it can be considered whether the energy demand for outdoor activities or for indoor activities will occur.

Furthermore, the grid independent input parameters 32 may include data about events, such as sporting events, trade fairs or concerts, which may also take into account a typical energy demand associated with these events.

Using the example of a dishwasher, the method is explained in more detail, it being apparent to one skilled in the art that the method is not limited to a dishwasher.

The control unit 2 of the dishwasher is set by the user a request 34, which may be, for example, that when switched on in the evening, the dishes should be washed well before breakfast. With this

Request, the requirement message 43 can be submitted to the higher-level rule element 2, from which also the room for maneuver results. If the utilization of the power supply network 1 is low, the dishwashing can be started immediately.

At a discounted night stream can also be provided that the

Dishwashing should take place in its tariff period. It can be due to the

Demand messages 43 of the control elements 2 of the energy demand to be determined during this time and react in good time to expected peak loads during a low-fare time. For example, it is expected that a variety of

Users will be able to observe a broadcast of a sports event on television during the low-fare tariff period

Transmission of sports event to be started.

The balancing between the control objects 2 takes place in each case via the hierarchically superior control object 2.

It may also be provided that only a certain number of dishwashers in a house flush at the same time to ensure a favorable load distribution. Furthermore, for example, the overall noise level can be kept low in order to keep the noise level low during the night hours.

The method may be readily provided with a computer program product, wherein the computer program product may be loaded directly into the internal memory of a computer and / or computer network and includes software code portions that perform the steps of the method described above when the computer program product is stored on a computer and / or a computer network is running.

For example, the software may be provided at least partially in the form of embedded software in one or more controllers.

Claims

PATENT APPLICATIONS

1. Method for controlling energy flows in one

Power supply network (1) having a plurality of hierarchically arranged units (91, 92, 93, 94, 95) connected by lines (8), the plurality of units (91, 92, 93, 94, 95) each having at least one control element (2). is associated, wherein the control elements (2) substantially according to the lines (8) of the associated units (91, 92, 93, 94, 95) coupled and in the region of the respectively associated unit (91, 92, 93, 94, 95) are arranged, wherein two mutually coupled control elements (2) are formed such that at least a part of the manipulated variable (41) of the hierarchically superior control element (2) as a target (31) of the hierarchically subordinate control element (2) is predetermined and the regulation is subsidiary ,

2. The method according to claim 1, characterized in that in at least one, in particular the plurality, the control elements (2) a machine learning model is used.

3. The method according to claim 2, characterized in that the machine-learning model is adaptively adapted to the power grid.

4. The method of claim 1, 2 or 3, characterized in that the control element (2) a target specification (31) is given that of the

Control element (2) energy supply network related input parameters (33) are obtained, and are determined by the control element (2) model parameters of the current power supply, and taking into account the

Model parameters and the target (31) manipulated variables (41) are determined.

5. The method according to claim 4, characterized in that the

Model parameters of a model device (51) of the control element (2) are determined, and that the manipulated variables (41) from a detection means (52) of the Control element (2) are determined.

6. The method according to claim 5, characterized in that by means of a judging means (53) a control strategy specification is determined and the manipulated variables (41) are determined by the determining means (52) taking into account the control strategy specification.

7. The method according to claim 5 or 6, characterized in that at least one, in particular the plurality, the control elements (2) both for the pattern device (51) a machine-leaming model is used as well as for the detection device (52) Machine learning model is used.

8. The method according to any one of claims 4 to 7, characterized in that the manipulated variables target specifications (31) for hierarchically subordinate

Control elements (2) and / or operating parameters (42) for the control element (2) associated unit (91, 92, 93, 94, 95) include.

9. The method according to any one of claims 5 to 8, characterized in that the model device (51) further model parameters of the future power supply are determined and the other model parameters for the determination of the manipulated variables (41) are used.

10. The method according to any one of claims 1 to 9, characterized in that in addition to the manipulated variable (41) a demand message (43) is determined and that the requirement message (43) is transmitted to the hierarchically superior control element (2).

11. The method according to claim 10, characterized in that from the requirement messages (43) of the hierarchically subordinate control elements (2) a request (34) as part of the energy supply network related

Input parameter (33) is formed.

12. The method according to any one of claims 1 to 11, characterized in that of the control element (2) power supply network independent

Input parameters (32) are related, and be taken into account in the determination of the manipulated variables (41).

13. The method according to any one of claims 1 to 12, characterized in that as a manipulated variable 41 at least one of the units 91, 92, 93, 94, 95 a

Pricing is selected.

14. Power grid control system for the regulation of

Energy flows in a power grid (1) with several means

Lines (8) connected hierarchically arranged units (91, 92, 93, 94, 95), wherein the plurality of units (91, 92, 93, 94, 95) are each assigned at least one control element (2), wherein the control elements ( 2) substantially corresponding to the lines (8) of the associated units (91, 92, 93, 94, 95) are coupled and in the region of the respectively associated unit (91, 92, 93, 94, 95) are arranged, wherein two coupled together Control elements (2) are designed such that the manipulated variable (41) of the hierarchically superior

Control element (2) a target (31) of the hierarchically subordinate

Rule element (2) and the regulation is subsidiary.

15. Power supply control system according to claim 14, characterized in that at least one, in particular the plurality, the

Control elements (2) includes a machine learning model.

16, power supply control system according to claim 14 or 15, characterized in that the control element (2) comprises a pattern device (51) and a detection means (52) and the model means (51) and the detection means (52) as a separate means coupled by means of interfaces modules are formed.

17. Power supply control system according to claim 16, characterized in that the control element (2) further a

Assessment device (53) and the assessment device (53) is designed as a separate module coupled by means of interfaces.

18. Power grid control system according to one of claims 14 to 17, characterized in that the power supply network (1) is an electrical power grid.

19. Power supply control system according to one of claims 14 to 18, characterized in that two mutually coupled control elements (2) are further formed such that a request message (43) of the hierarchically subordinate control element (2) to the hierarchically superior control element (2) is transmitted.

20. Computer program product directly into the internal memory of a

Computers and / or a computer network can be loaded and

Software code sections comprising the steps of the method according to one of claims 1 to 13 executed when the

Computer program product on a computer and / or one

Computer network is running.