CN116505594B - Method and system for determining adjustable droop coefficient of energy storage system based on error correction - Google Patents
Method and system for determining adjustable droop coefficient of energy storage system based on error correction Download PDFInfo
- Publication number
- CN116505594B CN116505594B CN202310764069.0A CN202310764069A CN116505594B CN 116505594 B CN116505594 B CN 116505594B CN 202310764069 A CN202310764069 A CN 202310764069A CN 116505594 B CN116505594 B CN 116505594B
- Authority
- CN
- China
- Prior art keywords
- power generation
- renewable energy
- generation equipment
- moment
- droop control
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000012937 correction Methods 0.000 title claims abstract description 51
- 238000004146 energy storage Methods 0.000 title claims abstract description 22
- 238000010248 power generation Methods 0.000 claims abstract description 109
- 238000007665 sagging Methods 0.000 claims abstract description 32
- 241000283153 Cetacea Species 0.000 claims description 69
- 238000013528 artificial neural network Methods 0.000 claims description 51
- 238000004422 calculation algorithm Methods 0.000 claims description 36
- 238000012549 training Methods 0.000 claims description 33
- 230000008569 process Effects 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 14
- 210000002569 neuron Anatomy 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 10
- 239000013598 vector Substances 0.000 claims description 10
- 238000012795 verification Methods 0.000 claims description 9
- 230000006978 adaptation Effects 0.000 claims description 8
- 230000003044 adaptive effect Effects 0.000 claims description 8
- 238000013178 mathematical model Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 6
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 7
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000004088 simulation Methods 0.000 description 6
- 238000011156 evaluation Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
- H02J3/466—Scheduling the operation of the generators, e.g. connecting or disconnecting generators to meet a given demand
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/381—Dispersed generators
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
- H02J2300/20—The dispersed energy generation being of renewable origin
Abstract
The application provides a method and a system for determining an adjustable droop coefficient of an energy storage system based on error correction, wherein the method comprises the following steps: acquiring wind speed information predicted values of renewable energy power generation equipment corresponding to a modularized multi-level converter in an MMC-MTDC system at each moment in a preset period; determining a predicted value of the generated power of the renewable energy power generation equipment at each moment; determining a droop control coefficient difference sequence corresponding to the renewable energy power generation equipment based on the initial droop control coefficient; determining a corrected droop control coefficient difference value corresponding to each time of the renewable energy power generation equipment; and determining the optimized sagging control coefficient corresponding to each moment of the renewable energy power generation equipment. According to the technical scheme provided by the application, the actual output condition of the renewable energy source is fully considered, and the sagging control coefficient is corrected by combining error correction, so that the output of the corresponding renewable energy source power generation equipment is more in line with the actual condition.
Description
Technical Field
The application relates to the field of sag control coefficient adjustment, in particular to a method and a system for determining an adjustable sag coefficient of an energy storage system based on error correction.
Background
The modular multilevel converter (Modular Multilevel Converter, MMC) is widely applied to a converter system due to good expansibility and output characteristics, compared with a traditional converter, the MMC not only effectively reduces voltage drop on a switching device and solves the problem of series voltage equalizing, but also has the characteristic of multilevel voltage regulation, and has the characteristics of low requirements of MMC on power electronic devices, simple networking mode and strong expansibility. The energy storage power station in the prior art has an energy storage type modularized multi-level converter using MMC, and sagging control can realize power distribution of the MMC under the condition of no high-speed communication, and has higher reliability and better flexibility, so that the energy storage power station is widely applied to the MMC. However, in the conventional voltage-power droop control, the droop control coefficient is a fixed value, the relationship between the direct current voltage and the active power can be represented by a straight line, the direct current voltage static difference adjustment can be realized through the active power adjustment, the direct current voltage stability of the MMC-MTDC (Multi-Terminal DirectCurrent based on Modular Multilevel Converter, multi-terminal direct current power transmission system based on the modularized Multi-level converter) is ensured, and the power distribution of each MMC (Modular Multi level Converter, modularized Multi-level converter) controller is realized; at present, droop control can be adjusted only according to a set droop control coefficient, flexibility is poor, so that each MMC submodule is difficult to balance between accurate power distribution and stable maintenance voltage, and the output of an MMC controller is not matched.
Disclosure of Invention
The application provides a method and a system for determining an adjustable droop coefficient of an energy storage system based on error correction, which at least solve the technical problems that droop control can be adjusted only according to a set droop control coefficient, flexibility is poor, so that each MMC submodule is difficult to balance between accurate power distribution and stable maintenance voltage, and the output of an MMC controller is not matched.
An embodiment of a first aspect of the present application provides a method for determining an adjustable sag coefficient of an energy storage system based on error correction, where the method includes:
acquiring wind speed information predicted values of renewable energy power generation equipment corresponding to a modularized multi-level converter in an MMC-MTDC system at each moment in a preset period;
inputting a wind speed information predicted value of the renewable energy power generation equipment corresponding to the modularized multi-level converter into a pre-trained power prediction model to obtain a power generation power predicted value of the renewable energy power generation equipment at each moment;
determining initial droop control coefficients corresponding to the renewable energy power generation equipment at each moment according to the power generation power predicted value at each moment, and determining droop control coefficient difference sequences corresponding to the renewable energy power generation equipment based on the initial droop control coefficients;
inputting the droop control coefficient difference value sequence into a pre-constructed error correction model, and carrying out optimization solution by utilizing an improved whale algorithm to obtain corrected droop control coefficient difference values corresponding to each moment of the renewable energy power generation equipment;
and determining the optimized droop control coefficient corresponding to each moment of the renewable energy power generation equipment based on the initial droop control coefficient corresponding to each moment of the renewable energy power generation equipment and the corrected droop control coefficient difference value.
Preferably, the pre-trained power prediction model is obtained by training an initial GRU neural network by taking a wind speed information predicted value of the renewable energy power generation equipment at each moment in a historical period as input and taking a generated power actual value at each moment as output;
the pre-constructed error correction model is obtained by training an initial BP neural network based on a difference sequence of droop control coefficients in a historical period and an improved whale algorithm.
Further, the obtaining process of the difference value sequence of the droop control coefficient in the history period includes:
and determining a difference value sequence of the droop control coefficients in the history period based on the droop control coefficients corresponding to the actual values of the generated power at each moment in the history period of the renewable energy power generation equipment and the droop control coefficients corresponding to the power set values corresponding to the MMC controller.
Further, the training process of the error correction model includes:
acquiring each difference sequence of the sagging control coefficients of the renewable energy power generation equipment in the history period, and the corrected sagging control coefficient difference sequence corresponding to each difference sequence;
constructing a training set and a verification set based on each difference sequence of the sagging control coefficients of the renewable energy power generation equipment in the history period and the corrected sagging control coefficient difference sequence corresponding to each difference sequence;
determining optimal parameters of an initial BP neural network by utilizing an improved whale algorithm;
and training the initial BP neural network based on the optimal parameters, the training set and the verification set to obtain a trained error correction model.
Further, the implementation steps of the improved whale algorithm include:
1) Surrounding target prey
Taking the position of the head whale closest to the position of the target object as an optimal solution, gradually updating the position of the head whale aiming at approaching the position of the optimal solution, wherein the calculation formula of a mathematical model surrounding the target object is as follows:
wherein D is the distance between the whale individual and the current optimal solution,is the globally optimal solution corresponding to the t-th iteration,for the position of whale corresponding to the t-th iteration, A is a first coefficient, C is a second coefficient,>for the position of whale corresponding to iteration t+1, wherein +.>、/>，/>For the first random vector, +.>For the second random vector,>、/>all belong to [0,1 ]]，/>，/>For astringing factor, ++>The maximum iteration number;
2) Spiral position update
And adjusting a position updating rule by adopting a mixed strategy combining an adaptive weight factor and random disturbance, wherein the calculation formula of a position updating model is as follows:
in the method, in the process of the application,is [0,1]First random number between->Is positioned at [ -1,1]First random number between->For the position of a random whale corresponding to the t-th iteration,>distance of random whale from the current optimal solution, < >>Is an adaptive weight factor;
3) Searching hunting object
Wherein, the calculation formula of the mathematical model of searching the prey stage is as follows:
in the method, in the process of the application,the positions of whale individuals randomly selected for the current iteration t.
Further, the determining optimal parameters of the initial BP neural network by using the improved whale algorithm comprises the following steps:
initializing adjustable parameters of a BP neural network, wherein an initial seed group rule n=20, a maximum iteration number lmax=50, and an independent variable number N=in+hn+hn+on+on to be optimized, wherein in is the number of neurons of an input layer of the BP neural network, hn is the number of neurons of an hidden layer of the BP neural network, and on is the number of neurons of an output layer of the BP neural network;
f2, taking the error of the training set as an adaptation function of an improved whale algorithm, and recording the current optimal adaptation value and the position information thereof;
f3, updating position information of whale bodies of the whale heads, judging whether the iteration number reaches the maximum iteration number, if so, ending the algorithm; if not, adding 1 to the current iteration number t, and returning to the step F2;
step F4: and outputting an optimal solution, and taking the weight and the threshold value obtained at the moment as optimal parameters of the BP neural network.
An embodiment of a second aspect of the present application provides a system for determining an adjustable sag coefficient of an energy storage system based on error correction, including:
the acquisition module is used for acquiring wind speed information predicted values of renewable energy power generation equipment corresponding to the modularized multi-level converter in the MMC-MTDC system at each moment in a preset period;
the first determining module is used for inputting the wind speed information predicted value of the renewable energy power generation equipment corresponding to the modularized multi-level converter into a pre-trained power prediction model to obtain the power generation power predicted value of the renewable energy power generation equipment at each moment;
the second determining module is used for determining initial droop control coefficients corresponding to the renewable energy power generation equipment at each moment according to the power generation power predicted value at each moment, and determining droop control coefficient difference value sequences corresponding to the renewable energy power generation equipment based on the initial droop control coefficients;
the correction module is used for inputting the sagging control coefficient difference value sequence into a pre-constructed error correction model, and carrying out optimization solution by utilizing an improved whale algorithm to obtain corrected sagging control coefficient difference values corresponding to each moment of the renewable energy power generation equipment;
and the third determining module is used for determining the optimized sagging control coefficient corresponding to each moment of the renewable energy power generation equipment based on the initial sagging control coefficient corresponding to each moment of the renewable energy power generation equipment and the corrected sagging control coefficient difference value.
Preferably, the pre-trained power prediction model is obtained by training an initial GRU neural network by taking a wind speed information predicted value of the renewable energy power generation equipment at each moment in a historical period as input and taking a generated power actual value at each moment as output;
the pre-constructed error correction model is obtained by training an initial BP neural network based on a difference sequence of droop control coefficients in a historical period and an improved whale algorithm.
An embodiment of a third aspect of the present application provides an electronic device, including: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method as described in the embodiments of the first aspect when the program is executed.
An embodiment of a fourth aspect of the present application proposes a computer readable storage medium, on which a computer program is stored, which program, when being executed by a processor, implements a method as described in the embodiment of the first aspect.
The technical scheme provided by the embodiment of the application at least has the following beneficial effects:
the application provides a method and a system for determining an adjustable droop coefficient of an energy storage system based on error correction, wherein the method comprises the following steps: acquiring wind speed information predicted values of renewable energy power generation equipment corresponding to a modularized multi-level converter in an MMC-MTDC system at each moment in a preset period; inputting a wind speed information predicted value of the renewable energy power generation equipment corresponding to the modularized multi-level converter into a pre-trained power prediction model to obtain a power generation power predicted value of the renewable energy power generation equipment at each moment; determining initial droop control coefficients corresponding to the renewable energy power generation equipment at each moment according to the power generation power predicted value at each moment, and determining droop control coefficient difference sequences corresponding to the renewable energy power generation equipment based on the initial droop control coefficients; inputting the droop control coefficient difference value sequence into a pre-constructed error correction model, and carrying out optimization solution by utilizing an improved whale algorithm to obtain corrected droop control coefficient difference values corresponding to each moment of the renewable energy power generation equipment; and determining the optimized droop control coefficient corresponding to each moment of the renewable energy power generation equipment based on the initial droop control coefficient corresponding to each moment of the renewable energy power generation equipment and the corrected droop control coefficient difference value. According to the technical scheme provided by the application, the actual output condition of the renewable energy source is fully considered, and the sagging control coefficient is corrected by combining error correction, so that the output of the corresponding renewable energy source power generation equipment is more in line with the actual condition.
Additional aspects and advantages of the application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the application.
Drawings
The foregoing and/or additional aspects and advantages of the application will become apparent and may be better understood from the following description of embodiments taken in conjunction with the accompanying drawings in which:
FIG. 1 is a flow chart of a method for determining an adjustable sag factor of an energy storage system based on error correction according to an embodiment of the present application;
FIG. 2 is a block diagram of a GRU neural network provided in accordance with one embodiment of the application;
FIG. 3 is a detailed schematic diagram of a method for determining an adjustable sag factor of an energy storage system based on error correction according to an embodiment of the present application;
FIG. 4 is a schematic diagram of simulation results using a fixed sag control factor according to one embodiment of the present application;
FIG. 5 is a schematic diagram of simulation results of an adjustable droop control coefficient using power prediction according to an embodiment of the present application;
fig. 6 is a block diagram of a system for determining an adjustable sag factor of an energy storage system based on error correction according to an embodiment of the present application.
Detailed Description
Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative and intended to explain the present application and should not be construed as limiting the application.
The application provides a method and a system for determining an adjustable droop coefficient of an energy storage system based on error correction, wherein the method comprises the following steps: acquiring wind speed information predicted values of renewable energy power generation equipment corresponding to a modularized multi-level converter in an MMC-MTDC system at each moment in a preset period; inputting a wind speed information predicted value of the renewable energy power generation equipment corresponding to the modularized multi-level converter into a pre-trained power prediction model to obtain a power generation power predicted value of the renewable energy power generation equipment at each moment; determining initial droop control coefficients corresponding to the renewable energy power generation equipment at each moment according to the power generation power predicted value at each moment, and determining droop control coefficient difference sequences corresponding to the renewable energy power generation equipment based on the initial droop control coefficients; inputting the droop control coefficient difference value sequence into a pre-constructed error correction model, and carrying out optimization solution by utilizing an improved whale algorithm to obtain corrected droop control coefficient difference values corresponding to each moment of the renewable energy power generation equipment; and determining the optimized droop control coefficient corresponding to each moment of the renewable energy power generation equipment based on the initial droop control coefficient corresponding to each moment of the renewable energy power generation equipment and the corrected droop control coefficient difference value. According to the technical scheme provided by the application, the actual output condition of the renewable energy source is fully considered, and the sagging control coefficient is corrected by combining error correction, so that the output of the corresponding renewable energy source power generation equipment is more in line with the actual condition.
The following describes a method and a system for determining an adjustable sag coefficient of an energy storage system based on error correction according to an embodiment of the present application with reference to the accompanying drawings.
Example 1
Fig. 1 is a flowchart of a method for determining an adjustable sag coefficient of an energy storage system based on error correction according to an embodiment of the present application, as shown in fig. 1, where the method includes:
step 1: and acquiring wind speed information predicted values of renewable energy power generation equipment corresponding to the modularized multi-level converter in the MMC-MTDC system at each moment in a preset period.
Step 2: and inputting the wind speed information predicted value of the renewable energy power generation equipment corresponding to the modularized multi-level converter into a pre-trained power prediction model to obtain the power generation power predicted value of the renewable energy power generation equipment at each moment.
In the embodiment of the disclosure, the pre-trained power prediction model is obtained by training an initial GRU neural network by taking a wind speed information predicted value of the renewable energy power generation equipment at each moment in a history period as input and taking a generated power actual value at each moment as output.
The structure of the GRU neural network is shown in fig. 2.
In the drawing the view of the figure,is the state of the previous moment relative to the current moment s;/>And->Input and output of GRU module at present moment respectively>And->For 2 key structures in the GRU module, namely a reset gate and an update gate, each gate is a simple neural network, and the activation function of the neural network adopts a sigmoid function in order to fix the output of the gate between 0 and 1>For the output candidate value after the reset gate processing, the structure of the GRU neural network is expressed by a formula as follows:
in the method, in the process of the application,for resetting the first parameter in the door, < +.>For resetting the second parameter in the door, < +.>For updating the first parameter in the gate, < >>For updating the second parameter in the gate, < >>To find the output candidate value +.>First parameter in the process,/->To find the output candidate value +.>A second parameter in the process, an operator). "means multiplying array elements in turn, +.>Reset gate for s time +.>Update gate for s time +.>Is in an internal state->For s time input, < >>Is the output candidate value at the time s,for the hidden layer state at s time, +.>For candidate hidden layer at s time, +.>Is the hidden layer state at s-1 time.
Step 3: and determining initial droop control coefficients corresponding to the renewable energy power generation equipment at each moment according to the power generation power predicted value at each moment, and determining a droop control coefficient difference sequence corresponding to the renewable energy power generation equipment based on the initial droop control coefficients.
Wherein, when the MMC-MTDC system is in a normal operation state, a DC voltage reference value is obtained according to the DC voltage droop control characteristicThen use the formula +.>Calculating a droop control coefficient corresponding to the MMC power set value and a droop control coefficient corresponding to the initial power predicted value, and obtaining a droop coefficient difference value by difference between the droop control coefficient and the droop control coefficient, wherein U is as follows 0 For the DC voltage set point, P, in the DC droop control curve 0 Setting value for MMC power; p is the actual power of the actual injected direct current power grid, namely the actual value of the generated power or the predicted value of the initial power, and m is the sagging control coefficient.
Step 4: and inputting the droop control coefficient difference value sequence into a pre-constructed error correction model, and carrying out optimization solution by utilizing an improved whale algorithm to obtain corrected droop control coefficient difference values corresponding to each time of the renewable energy power generation equipment.
In the embodiment of the disclosure, the pre-constructed error correction model is obtained by training an initial BP neural network based on a difference sequence of droop control coefficients in a history period and an improved whale algorithm;
the obtaining process of the difference value sequence of the droop control coefficient in the history period comprises the following steps:
and determining a difference value sequence of the droop control coefficients in the history period based on the droop control coefficients corresponding to the actual values of the generated power at each moment in the history period of the renewable energy power generation equipment and the droop control coefficients corresponding to the power set values corresponding to the MMC controller.
In an embodiment of the present disclosure, the training process of the error correction model includes:
acquiring each difference sequence of the sagging control coefficients of the renewable energy power generation equipment in the history period, and the corrected sagging control coefficient difference sequence corresponding to each difference sequence;
constructing a training set and a verification set based on each difference sequence of the sagging control coefficients of the renewable energy power generation equipment in the history period and the corrected sagging control coefficient difference sequence corresponding to each difference sequence;
determining optimal parameters of an initial BP neural network by utilizing an improved whale algorithm;
wherein the determining optimal parameters of the initial BP neural network using the modified whale algorithm comprises:
initializing adjustable parameters of a BP neural network, wherein an initial seed group rule n=20, a maximum iteration number lmax=50, and an independent variable number N=in+hn+hn+on+on to be optimized, wherein in is the number of neurons of an input layer of the BP neural network, hn is the number of neurons of an hidden layer of the BP neural network, and on is the number of neurons of an output layer of the BP neural network;
f2, taking the error of the training set as an adaptation function of an improved whale algorithm, and recording the current optimal adaptation value and the position information thereof;
f3, updating position information of whale bodies of the whale heads, judging whether the iteration number reaches the maximum iteration number, if so, ending the algorithm; if not, adding 1 to the current iteration number t, and returning to the step F2;
step F4: and outputting an optimal solution, and taking the weight and the threshold value obtained at the moment as optimal parameters of the BP neural network.
Further, the implementation steps of the improved whale algorithm include:
1) Surrounding target prey
Taking the position of the head whale closest to the position of the target object as an optimal solution, gradually updating the position of the head whale aiming at approaching the position of the optimal solution, wherein the calculation formula of a mathematical model surrounding the target object is as follows:
wherein D is the distance between the whale individual and the current optimal solution,is the t timeThe corresponding globally optimal solution is iterated,for the position of whale corresponding to the t-th iteration, A is a first coefficient, C is a second coefficient,>for the position of whale corresponding to iteration t+1, wherein +.>、/>，/>For the first random vector, +.>As a second random vector of the set of random vectors,、/>all belong to [0,1 ]]，/>，/>For astringing factor, ++>The maximum iteration number;
2) Spiral position update
And adjusting a position updating rule by adopting a mixed strategy combining an adaptive weight factor and random disturbance, wherein the calculation formula of a position updating model is as follows:
in the method, in the process of the application,is [0,1]First random number between->Is positioned at [ -1,1]First random number between->For the position of a random whale corresponding to the t-th iteration,>distance of random whale from the current optimal solution, < >>Is an adaptive weight factor;
3) Searching hunting object
Wherein, the calculation formula of the mathematical model of searching the prey stage is as follows:
in the method, in the process of the application,the positions of whale individuals randomly selected for the current iteration t.
And training the initial BP neural network based on the optimal parameters, the training set and the verification set to obtain a trained error correction model.
It should be noted that, the BP neural network is a multi-layer feedforward network trained by an error back propagation algorithm, and includes an input layer, an hidden layer, and an output layer, and is divided into two stages of forward transmission and back propagation, in which a signal in the forward transmission is transmitted from the input layer to the output layer through the hidden layer through a series of calculations to calculate a predicted value, and if the error between the predicted value and an actual value does not reach a threshold value, the weight and the bias of the network are updated through error back propagation, so as to reduce the error.
Step 5: and determining the optimized droop control coefficient corresponding to each moment of the renewable energy power generation equipment based on the initial droop control coefficient corresponding to each moment of the renewable energy power generation equipment and the corrected droop control coefficient difference value.
For example, the initial droop control coefficient and the corrected droop control coefficient difference are added to obtain the optimized droop control coefficient.
It should be noted that, the detailed flow of the method according to this embodiment may be shown in fig. 3.
In order to accurately compare the performances of each model under various prediction conditions, average absolute error (Mean absolute error, MAE), average absolute percent error (Mean absolute percentage error, MAPE) and root mean square error (Root mean square error, RMSE) are selected as model evaluation standards, and the calculation formulas are respectively shown in formulas (1), (2) and (3), wherein all models use MAE as a loss function during training.
Wherein, the liquid crystal display device comprises a liquid crystal display device,（1），/>（2）、/>(3) Wherein n is the number of samples, +.>、/>I is the predicted value of the sagging control coefficient at the moment and i is the actual value of the sagging control coefficient, respectively,/->Is the mean absolute error value, +.>Is flatAbsolute percentage error value->The model evaluation criteria are compared for root mean square error values, as shown in table 1;
TABLE 1
The single-step prediction and multi-step prediction evaluation criteria of each model are shown in the table. The model GRU-IWOA-BP MAE, MAPE, RMSE provided by the embodiment is lowest and highest in accuracy.
Wherein MAPE is used as a main evaluation index, and the performances of each model are analyzed to obtain the following conclusion:
compared with BP and GRU, WOA-BP obviously improves the prediction precision, MAPE is respectively reduced by 3.50%, 53.14%,44.67% and 3.33, and the WOA algorithm is proved to optimize the initial weight threshold of the BP neural network, so that a stable WOA-BP prediction model is established.
Compared with WOA-BP, the IWA-BP obviously improves the prediction precision, MAPE is respectively reduced by 49.50 percent and 42.09 percent, and the IOWA is proved to improve the local searching capability of WOA.
Compared with WOA-BP, the IWA-BP obviously improves the prediction precision, MAPE is respectively reduced by 49.50 percent and 42.09 percent, and the IOWA is proved to improve the local searching capability of WOA.
Compared with the IWOA-BP, the MAPE of the GRU-IWOA-BP of the proposed model is respectively reduced by 38.54 percent and 33.44 percent, which shows that an error correction model in the proposed model can effectively extract the hidden information of an error sequence generated by a preliminary prediction model, correct the preliminary prediction error and improve the overall prediction precision.
Further, the two inverters have unequal capacities and are subjected to verification of a droop control strategy: and the capacity of the active power is 2 times that of the inverter 2 by using the inverter 1. Simulation time was set to 2.5s, lda1=100000 kW was added at t=1.0s, lda2=200000 kW was added at t=2.0s, and simulation with a fixed droop control and simulation with an adjustable droop control for power prediction were performed, with simulation results shown in fig. 4, fig. 5 and table 2:
TABLE 2
It can be seen from fig. 4 that the total power output after the load is added is satisfactory, but the active power cannot be output in proportion to the capacity. Fig. 5 shows that the total power output after the load is added meets the requirement, and the active power is output according to the capacity proportion.
As can be taken from table 2, load1 was added at t=0.1 s, and the voltage was reduced from 325V to 314V; load2 was added at t=2s and the voltage dropped from 314V to 295V. As can be taken from table 2, load1 was added at t=1s, and the voltage was reduced from 325V to 316V; by adding at t2s, the voltage can be restored to normal value. Therefore, the droop control based on error correction can realize the balance control of power balance and voltage stabilization of the inverter.
In summary, according to the method for determining the adjustable droop coefficient of the energy storage system based on the error correction provided by the embodiment, the actual output condition of the renewable energy source is fully considered, and the droop control coefficient is corrected by combining the error correction, so that the output of the corresponding renewable energy source power generation equipment is more in line with the actual condition.
Example two
Fig. 6 is a block diagram of a system for determining an adjustable sag factor of an energy storage system based on error correction according to an embodiment of the present application, as shown in fig. 6, where the system includes:
the obtaining module 100 is configured to obtain wind speed information predicted values of renewable energy power generation devices corresponding to the modularized multi-level converter in the MMC-MTDC system at each moment in a preset period;
the first determining module 200 is configured to input a wind speed information predicted value of the renewable energy power generation device corresponding to the modular multilevel converter into a pre-trained power prediction model, so as to obtain a power generation power predicted value of the renewable energy power generation device at each moment;
a second determining module 300, configured to determine an initial droop control coefficient corresponding to each time of the renewable energy power generation device according to the predicted value of the generated power at each time, and determine a droop control coefficient difference sequence corresponding to the renewable energy power generation device based on the initial droop control coefficient;
the correction module 400 is configured to input the droop control coefficient difference value sequence into a pre-constructed error correction model, and perform optimization solution by using an improved whale algorithm to obtain corrected droop control coefficient difference values corresponding to each time of the renewable energy power generation device;
and a third determining module 500, configured to determine an optimized droop control coefficient corresponding to each time of the renewable energy power generation device based on the initial droop control coefficient corresponding to each time of the renewable energy power generation device and the corrected droop control coefficient difference value.
In the embodiment of the disclosure, the pre-trained power prediction model is obtained by training an initial GRU neural network by taking a wind speed information predicted value of the renewable energy power generation equipment at each moment in a history period as input and taking a generated power actual value at each moment as output;
the pre-constructed error correction model is obtained by training an initial BP neural network based on a difference sequence of droop control coefficients in a historical period and an improved whale algorithm.
Further, the obtaining process of the difference value sequence of the droop control coefficient in the history period includes:
and determining a difference value sequence of the droop control coefficients in the history period based on the droop control coefficients corresponding to the actual values of the generated power at each moment in the history period of the renewable energy power generation equipment and the droop control coefficients corresponding to the power set values corresponding to the MMC controller.
Further, the training process of the error correction model includes:
acquiring each difference sequence of the sagging control coefficients of the renewable energy power generation equipment in the history period, and the corrected sagging control coefficient difference sequence corresponding to each difference sequence;
constructing a training set and a verification set based on each difference sequence of the sagging control coefficients of the renewable energy power generation equipment in the history period and the corrected sagging control coefficient difference sequence corresponding to each difference sequence;
determining optimal parameters of an initial BP neural network by utilizing an improved whale algorithm;
and training the initial BP neural network based on the optimal parameters, the training set and the verification set to obtain a trained error correction model.
It should be noted that the implementation steps of the improved whale algorithm include:
1) Surrounding target prey
Taking the position of the head whale closest to the position of the target object as an optimal solution, gradually updating the position of the head whale aiming at approaching the position of the optimal solution, wherein the calculation formula of a mathematical model surrounding the target object is as follows:
wherein D is the distance between the whale individual and the current optimal solution,is the globally optimal solution corresponding to the t-th iteration,for the position of whale corresponding to the t-th iteration, A is a first coefficient, C is a second coefficient,>for the position of whale corresponding to iteration t+1, wherein +.>、/>，/>For the first random vector, +.>As a second random vector of the set of random vectors,、/>all belong to [0,1 ]]，/>，/>For astringing factor, ++>The maximum iteration number;
2) Spiral position update
And adjusting a position updating rule by adopting a mixed strategy combining an adaptive weight factor and random disturbance, wherein the calculation formula of a position updating model is as follows:
in the method, in the process of the application,is [0,1]First random number between->Is positioned at [ -1,1]First random number between->For the position of a random whale corresponding to the t-th iteration,>distance of random whale from the current optimal solution, < >>Is an adaptive weight factor;
3) Searching hunting object
Wherein, the calculation formula of the mathematical model of searching the prey stage is as follows:
in the method, in the process of the application,the positions of whale individuals randomly selected for the current iteration t.
Wherein the determining optimal parameters of the initial BP neural network using the modified whale algorithm comprises:
step E1: initializing adjustable parameters of the BP neural network, wherein the initial seed group rule n=20, the maximum iteration number lmax=50, and the number of independent variables N=in+hn+hn+on+on to be optimized, wherein in is the number of neurons of an input layer of the BP neural network, hn is the number of neurons of an hidden layer of the BP neural network, and on is the number of neurons of an output layer of the BP neural network;
e2, taking the error of the training set as an adaptation function of an improved whale algorithm, and recording the current optimal adaptation value and the position information thereof;
e3, updating position information of whale individuals, judging whether the iteration number reaches the maximum iteration number, if so, ending the algorithm; if not, adding 1 to the current iteration number t, and returning to the step E2;
step E4: and outputting an optimal solution, and taking the weight and the threshold value obtained at the moment as optimal parameters of the BP neural network.
In summary, the determining system for the adjustable droop coefficient of the energy storage system based on the error correction provided by the embodiment fully considers the actual output condition of the renewable energy source, and corrects the droop control coefficient by combining the error correction, so that the output of the corresponding renewable energy source power generation equipment is more in line with the actual condition.
Example III
In order to achieve the above embodiments, the present disclosure further proposes an electronic device including: a memory, a processor, and a computer program stored on the memory and executable on the processor, which when executed, implements the method as described in embodiment one.
Example IV
In order to implement the above-described embodiments, the present disclosure also proposes a computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, implements a method as described in embodiment one.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and additional implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order from that shown or discussed, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present application.
While embodiments of the present application have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the application, and that variations, modifications, alternatives and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the application.
Claims (8)
1. A method for determining an adjustable sag factor of an energy storage system based on error correction, the method comprising:
acquiring wind speed information predicted values of renewable energy power generation equipment corresponding to a modularized multi-level converter in an MMC-MTDC system at each moment in a preset period;
inputting wind speed information predicted values of renewable energy power generation equipment corresponding to the modularized multi-level converter into a pre-trained power prediction model to obtain power generation power predicted values of the renewable energy power generation equipment at all times, wherein the pre-trained power prediction model is obtained by training an initial GRU neural network by taking the wind speed information predicted values of the renewable energy power generation equipment at all times in a historical period as input and taking the actual power generation power values at all times as output;
determining initial droop control coefficients corresponding to the renewable energy power generation equipment at each moment according to the power generation power predicted value at each moment, and determining droop control coefficient difference sequences corresponding to the renewable energy power generation equipment based on the initial droop control coefficients;
inputting the droop control coefficient difference value sequence into a pre-constructed error correction model, and carrying out optimization solving by utilizing an improved whale algorithm to obtain corrected droop control coefficient difference values corresponding to each moment of the renewable energy power generation equipment, wherein the pre-constructed error correction model is obtained by training an initial BP neural network based on the droop control coefficient difference value sequence in a historical period and the improved whale algorithm;
and determining the optimized droop control coefficient corresponding to each moment of the renewable energy power generation equipment based on the initial droop control coefficient corresponding to each moment of the renewable energy power generation equipment and the corrected droop control coefficient difference value.
2. The method of claim 1, wherein the step of obtaining the sequence of differences in droop control coefficients over the history period comprises:
and determining a difference value sequence of the droop control coefficients in the history period based on the droop control coefficients corresponding to the actual values of the generated power at each moment in the history period of the renewable energy power generation equipment and the droop control coefficients corresponding to the power set values corresponding to the MMC controller.
3. The method of claim 2, wherein the training process of the error correction model comprises:
acquiring each difference sequence of the sagging control coefficients of the renewable energy power generation equipment in the history period, and the corrected sagging control coefficient difference sequence corresponding to each difference sequence;
constructing a training set and a verification set based on each difference sequence of the sagging control coefficients of the renewable energy power generation equipment in the history period and the corrected sagging control coefficient difference sequence corresponding to each difference sequence;
determining optimal parameters of an initial BP neural network by utilizing an improved whale algorithm;
and training the initial BP neural network based on the optimal parameters, the training set and the verification set to obtain a trained error correction model.
4. A method as claimed in claim 3, wherein the step of implementing the modified whale algorithm comprises:
1) Surrounding target prey
Taking the position of the head whale closest to the position of the target object as an optimal solution, gradually updating the position of the head whale aiming at approaching the position of the optimal solution, wherein the calculation formula of a mathematical model surrounding the target object is as follows:
wherein D is whale individual and currentThe distance of the front optimal solution,is the globally optimal solution corresponding to the t-th iteration,>for the position of whale corresponding to the t-th iteration, A is a first coefficient, C is a second coefficient,>for the position of whale corresponding to iteration t+1, wherein +.>、/>，/>For the first random vector, +.>For the second random vector,>、/>all belong to [0,1 ]]，/>，/>For astringing factor, ++>The maximum iteration number;
2) Spiral position update
And adjusting a position updating rule by adopting a mixed strategy combining an adaptive weight factor and random disturbance, wherein the calculation formula of a position updating model is as follows:
in the method, in the process of the application,is [0,1]First random number between->Is positioned at [ -1,1]First random number between->For the position of a random whale corresponding to the t-th iteration,>distance of random whale from the current optimal solution, < >>Is an adaptive weight factor;
3) Searching hunting object
Wherein, the calculation formula of the mathematical model of searching the prey stage is as follows:
in the method, in the process of the application,the positions of whale individuals randomly selected for the current iteration t.
5. The method of claim 4, wherein said determining optimal parameters for an initial BP neural network using a modified whale algorithm comprises:
initializing adjustable parameters of a BP neural network, wherein an initial seed group rule n=20, a maximum iteration number lmax=50, and an independent variable number N=in+hn+hn+on+on to be optimized, wherein in is the number of neurons of an input layer of the BP neural network, hn is the number of neurons of an hidden layer of the BP neural network, and on is the number of neurons of an output layer of the BP neural network;
f2, taking the error of the training set as an adaptation function of an improved whale algorithm, and recording the current optimal adaptation value and the position information thereof;
f3, updating position information of whale bodies of the whale heads, judging whether the iteration number reaches the maximum iteration number, if so, ending the algorithm; if not, adding 1 to the current iteration number t, and returning to the step F2;
step F4: and outputting an optimal solution, and taking the weight and the threshold value obtained at the moment as optimal parameters of the BP neural network.
6. A system for determining an adjustable sag factor of an energy storage system based on error correction, the system comprising:
the acquisition module is used for acquiring wind speed information predicted values of renewable energy power generation equipment corresponding to the modularized multi-level converter in the MMC-MTDC system at each moment in a preset period;
the first determining module is used for inputting the wind speed information predicted value of the renewable energy power generation equipment corresponding to the modularized multi-level converter into a pre-trained power prediction model to obtain the power generation power predicted value of the renewable energy power generation equipment at each moment, wherein the pre-trained power prediction model is obtained by training an initial GRU neural network by taking the wind speed information predicted value of the renewable energy power generation equipment at each moment in a historical period as input and taking the actual power generation power value at each moment as output;
the second determining module is used for determining initial droop control coefficients corresponding to the renewable energy power generation equipment at each moment according to the power generation power predicted value at each moment, and determining droop control coefficient difference value sequences corresponding to the renewable energy power generation equipment based on the initial droop control coefficients;
the correction module is used for inputting the droop control coefficient difference value sequence into a pre-built error correction model, and carrying out optimization solving by utilizing an improved whale algorithm to obtain corrected droop control coefficient difference values corresponding to each time of the renewable energy power generation equipment, wherein the pre-built error correction model is obtained by training an initial BP neural network based on the droop control coefficient difference value sequence in a historical period and the improved whale algorithm;
and the third determining module is used for determining the optimized sagging control coefficient corresponding to each moment of the renewable energy power generation equipment based on the initial sagging control coefficient corresponding to each moment of the renewable energy power generation equipment and the corrected sagging control coefficient difference value.
7. An electronic device, comprising: memory, a processor and a computer program stored on the memory and executable on the processor, which processor implements the method according to any of claims 1-5 when executing the program.
8. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the method according to any of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310764069.0A CN116505594B (en) | 2023-06-27 | 2023-06-27 | Method and system for determining adjustable droop coefficient of energy storage system based on error correction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310764069.0A CN116505594B (en) | 2023-06-27 | 2023-06-27 | Method and system for determining adjustable droop coefficient of energy storage system based on error correction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116505594A CN116505594A (en) | 2023-07-28 |
| CN116505594B true CN116505594B (en) | 2023-10-03 |
Family
ID=87316963
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310764069.0A Active CN116505594B (en) | 2023-06-27 | 2023-06-27 | Method and system for determining adjustable droop coefficient of energy storage system based on error correction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116505594B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5664889B1 (en) * | 2014-03-06 | 2015-02-04 | 全北大学校 産学協力▲団▼ | Inertia control method of wind power generator based on time-varying droop |
| CN114430165A (en) * | 2021-11-25 | 2022-05-03 | 南京师范大学 | Micro-grid group intelligent coordination control method and device based on depth model prediction |
| CN114915225A (en) * | 2022-04-22 | 2022-08-16 | 昆明理工大学 | Permanent magnet synchronous motor parameter identification method based on optimized BP neural network |
| CN115049041A (en) * | 2022-04-15 | 2022-09-13 | 西安石油大学 | Moisture pipe liquid holding rate prediction method based on WOA-BP neural network |
| CN115663920A (en) * | 2022-11-09 | 2023-01-31 | 山东电力工程咨询院有限公司 | Method and system for adjusting adjustable droop coefficient based on power predicted value |
-
2023
- 2023-06-27 CN CN202310764069.0A patent/CN116505594B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5664889B1 (en) * | 2014-03-06 | 2015-02-04 | 全北大学校 産学協力▲団▼ | Inertia control method of wind power generator based on time-varying droop |
| CN114430165A (en) * | 2021-11-25 | 2022-05-03 | 南京师范大学 | Micro-grid group intelligent coordination control method and device based on depth model prediction |
| CN115049041A (en) * | 2022-04-15 | 2022-09-13 | 西安石油大学 | Moisture pipe liquid holding rate prediction method based on WOA-BP neural network |
| CN114915225A (en) * | 2022-04-22 | 2022-08-16 | 昆明理工大学 | Permanent magnet synchronous motor parameter identification method based on optimized BP neural network |
| CN115663920A (en) * | 2022-11-09 | 2023-01-31 | 山东电力工程咨询院有限公司 | Method and system for adjusting adjustable droop coefficient based on power predicted value |
Non-Patent Citations (3)
| Title |
|---|
| A Hybrid Forecasting Model for Wind Energy based on the Complementary Ensemble Empirical Mode Decomposition and Whale Optimized Back Propagation Neural Network;Weihang Li et al.;《2020 IEEE 4th Conference on Energy Internet and Energy System Integration (EI2)》;第1084-1089页 * |
| 基于鲸鱼算法改进下垂控制的微电网平滑切换策略;肖涵麒等;《北华大学学报》;第24卷(第1期);第128-133页 * |
| 改进鲸鱼算法及其在浅层神经网络搜索中的权值阈值优化;刘 威等;《控制与决策》;第38卷(第4期);第1144-1152页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116505594A (en) | 2023-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107437813B (en) | Power distribution network reactive power optimization method based on cuckoo-particle swarm | |
| CN114362187B (en) | Active power distribution network cooperative voltage regulation method and system based on multi-agent deep reinforcement learning | |
| CN111367349A (en) | Photovoltaic MPPT control method and system based on prediction model | |
| CN112803434A (en) | Reactive power optimization method, device, equipment and storage medium for active power distribution network | |
| CN111682557B (en) | Optimization method for configuration place and optimal compensation capacity of reactive power equipment of power system | |
| CN112330487B (en) | Photovoltaic power generation short-term power prediction method | |
| CN110675278A (en) | Photovoltaic power short-term prediction method based on RBF neural network | |
| CN112953286A (en) | Q-Learning algorithm-based parallel circulating current suppression method for same-capacity three-phase inverter | |
| CN108448659B (en) | Method for solving reactive power optimization of power system based on HACS algorithm | |
| CN113780664A (en) | Time sequence prediction method based on TDT-SSA-BP | |
| Gholizadeh et al. | A comparative study of reinforcement learning algorithms for distribution network reconfiguration with deep Q-learning-based action sampling | |
| CN116505594B (en) | Method and system for determining adjustable droop coefficient of energy storage system based on error correction | |
| CN114722693A (en) | Optimization method of two-type fuzzy control parameter of water turbine regulating system | |
| CN107871157B (en) | Data prediction method, system and related device based on BP and PSO | |
| CN116760047A (en) | Power distribution network voltage reactive power control method and system based on safety reinforcement learning algorithm | |
| CN112346342B (en) | Single-network self-adaptive evaluation design method of non-affine dynamic system | |
| Sefriti et al. | Neural network Incremental conductance MPPT algorithm for photovoltaic water pumping system | |
| CN112152267B (en) | Power grid random reactive power optimization scheduling method considering source load uncertainty | |
| CN112182948A (en) | Farmland multi-target control drainage model solving method based on vector angle particle swarm | |
| Wongsathan | Performance of hybrid fuzzy and neuro-fuzzy controller with MOHGA based MPPT for a Solar PV module on various weather conditions | |
| Ni et al. | LSHADE with semi-parameter adaptation for chaotic time series prediction | |
| CN111859780A (en) | Micro-grid operation optimization method and system | |
| Pei et al. | Multi-agent deep reinforcement learning for realistic distribution system voltage control using pv inverters | |
| CN116581766B (en) | Virtual power plant strengthening online voltage control method considering sagging characteristic | |
| Iliya et al. | Differential Evolution Based PID Antenna Position Control System |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |