Emergency power control-oriented micro-grid regulation and control threshold intelligent setting method

Abstract

The invention discloses a micro-grid regulation and control threshold intelligent setting method for emergency power control. The method comprises the following steps: step 1, constructing a micro-grid emergency power control scene set; 2, a generative adversarial network model is constructed, the generative adversarial network model comprises an emergency power control scene generator and a microgrid regulation and control threshold effect discriminator, and the microgrid emergency power control scene set is input into the generative adversarial network model for multiple rounds of adversarial learning, parameters of a self reward function are optimized and solved through an improved plant cell group algorithm of the emergency power control scene generator, so that a current micro-grid regulation and control threshold value is obtained, and whether the current micro-grid regulation and control threshold value meets the micro-grid emergency power regulation and control requirement or not is judged by a micro-grid regulation and control threshold value effect discriminator. If not, iteration adversarial learning continues, and if yes, iteration is stopped, and the current micro-grid regulation and control threshold value is output. According to the invention, the reliability and stability of the micro-grid in a complex power system can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of microgrid control methods, and in particular to an intelligent setting method for microgrid control thresholds for emergency power control. Background Art

[0002] With the development of new power systems and the large-scale access of renewable energy, microgrids are becoming increasingly important in modern power systems. Microgrids can not only improve power quality and power supply reliability, but also provide rapid and accurate responses in emergency situations to ensure the stable operation of power systems. However, the setting of control thresholds for microgrids in emergency power control is a key issue. Appropriate control thresholds can not only improve the operating efficiency of microgrids, but also effectively avoid false operations and ensure the safety and stability of the system. Therefore, studying a setting method for intelligent control thresholds has important practical significance and academic value.

[0003] Although existing research has made some progress in the setting of microgrid control thresholds, there are still many limitations. Most traditional methods rely on fixed thresholds or empirical values, which are difficult to adapt to the needs of different operating conditions, resulting in the microgrid's response speed and control accuracy being limited in emergency situations, and may even cause malfunctions and affect the stability of the system, which is inadequate when dealing with the complex and changeable power system operating environment. In addition, when dealing with the problem of microgrid control thresholds, existing optimization algorithms often have problems such as low search efficiency and easy to fall into local optimality, making it difficult to meet the requirements of high precision and rapid response. These limitations significantly affect the response speed and control accuracy of microgrids in emergency situations, and restrict their application effects. Summary of the invention

[0004] The present invention provides a microgrid control threshold intelligent setting method for emergency power control, so as to solve the problems of low efficiency and easy falling into local optimum when the existing optimization algorithm in the prior art processes the microgrid control threshold.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] A microgrid control threshold intelligent setting method for emergency power control, comprising the following steps:

[0007] Step 1: Obtain the grid emergency power regulation demand information, microgrid output data, and microgrid status information, and construct a microgrid emergency power control scenario set based on the grid emergency power regulation demand information, microgrid output data, and microgrid status information;

[0008] Step 2: construct a generative adversarial network model for microgrid regulation threshold setting, wherein the generative adversarial network model includes an emergency power control scenario generator and a microgrid regulation threshold effect discriminator, and the microgrid emergency power control scenario set obtained in step 1 is input into the emergency power control scenario generator in the generative adversarial network model to perform multi-round adversarial learning of the generative adversarial network model;

[0009] In each round of adversarial learning, the emergency power control scenario generator optimizes and solves the parameters of its own reward function based on the improved plant cell swarm algorithm, and generates the current microgrid emergency power control scenario and the current microgrid regulation threshold corresponding to the current microgrid emergency power control scenario under the guidance of the optimized reward function, and the microgrid regulation threshold effect discriminator determines whether the current microgrid regulation threshold meets the microgrid emergency power regulation demand;

[0010] If the judgment result of the microgrid regulation threshold effect discriminator is that the current microgrid regulation threshold cannot meet the microgrid emergency power regulation demand, the microgrid regulation threshold effect discriminator feeds back the judgment result to the emergency power control scenario generator, and generates an adversarial network model to repeat the above process for the next round of adversarial learning;

[0011] If the judgment result is that the current microgrid control threshold can meet the microgrid emergency power control demand, the generative adversarial network model stops adversarial learning, and the emergency power control scenario generator outputs the current microgrid control threshold, thereby completing the microgrid control threshold setting.

[0012] Furthermore, in step 1, the microgrid output data includes microgrid photovoltaic output data and microgrid wind power output data.

[0013] Furthermore, in step 1, the microgrid status information includes the microgrid energy storage device status, the microgrid load status, and the microgrid adjustable resource status.

[0014] Furthermore, in step 2, after the noise is centrally superimposed on the microgrid emergency power control scenario, it is input into the emergency power control scenario generator to simulate the uncertainty of microgrid operation by superimposing the noise.

[0015] Furthermore, in step 2, the emergency power control scenario generator is composed of multiple layers of fully connected layers and a universal activation function. When the microgrid emergency power control scenario set is input into the emergency power control scenario generator, each fully connected layer of the emergency power control scenario generator extracts the features of the microgrid emergency power control scenario set (including scenario status and threshold), and uses a universal activation function to determine whether the extracted features are transmitted to the next fully connected layer; if the judgment result is yes, the features extracted by the current fully connected layer can be transmitted to the next fully connected layer; the transmission is performed successively until the extracted microgrid emergency power control scenario set features are transmitted to the end of the emergency power control scenario generator to obtain the current emergency power control scenario set status and threshold.

[0016] Furthermore, in step 2, the microgrid regulation threshold effect discriminator is composed of multiple layers of fully connected layers and a universal activation function. When the current emergency power control scenario set state and threshold generated by the emergency power control scenario generator are input into the microgrid regulation threshold effect discriminator, each fully connected layer of the microgrid regulation threshold effect discriminator extracts the current emergency power control scenario set state and the microgrid operation safety state under the threshold, and uses a universal activation function to determine whether the extracted features are transmitted to the next fully connected layer; if the judgment result is yes, the features extracted by the current fully connected layer can be transmitted to the next fully connected layer; the features are transmitted successively until the extracted microgrid operation safety state features are transmitted to the end of the microgrid regulation threshold effect discriminator to obtain the current microgrid operation safety state features, and then determine whether the current microgrid operation safety state features meet the microgrid operation requirements.

[0017] Furthermore, in step 2, the current microgrid control threshold obtained by the emergency power control scenario generator includes a control power shortage threshold and a control delay time.

[0018] Furthermore, in step 2, the reward function of the emergency power control scenario generator is as follows:

[0019] R(λ)=ω(λ n -λ n-1 )+b,2≤n≤N

[0020] Where: R(λ) is the reward function of the emergency power control scenario generator; n is the sample generated by the emergency power control scenario generator for the nth time; ω is the sensitivity coefficient of the emergency power control scenario generator. b is the bias coefficient of the emergency power control scenario generator; N is the total number of samples generated by the emergency power control scenario generator;

[0021] In each round of adversarial learning, the emergency power control scenario generator optimizes the sensitivity coefficient ω and the bias coefficient b in the reward function based on the improved plant cell swarm algorithm.

[0022] The present invention proposes an intelligent tuning method combining a generative adversarial network and an improved plant cell swarm algorithm. By combining the generative adversarial network and the improved plant cell swarm algorithm, the algorithm parameters can be dynamically adjusted at different stages to improve the search efficiency and avoid falling into the local optimum, thereby achieving the optimal setting of the control threshold.

[0023] The present invention solves the limitations of traditional methods when dealing with complex and changeable power system operating environments. By adjusting the dynamic learning factor, the performance of the generative adversarial network at different stages is improved, and the improved plant cell swarm algorithm is used to perform secondary optimization on the generative adversarial network to take into account the global exploration ability and local mining ability of the algorithm. In this way, it can not only significantly improve the setting accuracy of the control threshold, improve the response speed and control accuracy of the microgrid, ensure the rapid response and efficient operation of the microgrid in emergency situations, but also adapt to the needs of different operating conditions, ensure the stable operation of the system, and avoid misoperation. Through the application of this method, the microgrid control threshold can be quickly set, which can effectively improve the reliability and stability of the microgrid in a complex power system, and provide solid technical support for the intelligent development of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of the method of an embodiment of the present invention.

[0025] Figure 2 It is a simulation analysis result diagram of the performance of three different methods in the experiment of the embodiment of the present invention.

[0026] Figure 3 It is a simulation analysis result diagram of the optimization performance comparison of the improved plant cell colony algorithm in the experiment of the embodiment of the present invention. DETAILED DESCRIPTION

[0027] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0028] like Figure 1 As shown, this embodiment discloses a microgrid control threshold intelligent setting method for emergency power control, comprising the following steps:

[0029] Step 1: Construct a microgrid emergency power control scenario set

[0030] In this embodiment, the grid emergency power regulation demand information, microgrid output data, and microgrid status information are obtained from the microgrid, wherein the microgrid output data includes the microgrid photovoltaic output data and the microgrid wind power output data, and the microgrid status information includes the microgrid energy storage device status, the microgrid load status, and the microgrid adjustable resource status. Thus, the microgrid emergency power control scenario set is constructed based on the grid emergency power regulation demand information, microgrid photovoltaic output data, microgrid wind power output data, microgrid energy storage device status, microgrid load status, and microgrid adjustable resource status.

[0031] Specifically, the grid emergency power regulation demand information is a time series power demand vector, as shown in formula (1):

[0032]

[0033] In formula (1), is the sequential power demand vector; They are the power demands at time 1, 2, …n respectively.

[0034] The photovoltaic output data of the microgrid can be expressed as shown in formula (2):

[0035]

[0036] In formula (2), Y PV It is the photovoltaic output data set; They are the photovoltaic output data at the 1st, 2nd…nth moments respectively.

[0037] The wind power output data of the microgrid can be expressed as shown in formula (3):

[0038]

[0039] In formula (3), Y W It is a wind power output dataset; They are the wind power output data at the 1st, 2nd…nth moments respectively.

[0040] The state of the microgrid energy storage device can be expressed as shown in formula (4):

[0041]

[0042] In formula (4), Y ES is the state vector of the microgrid energy storage device; They are the charge states of the microgrid energy storage devices collected at the 1st, 2nd…nth moments respectively.

[0043] The load state of the microgrid can be expressed as shown in formula (5):

[0044]

[0045] In formula (5), Y LOAD is the microgrid load state time series vector; They are the microgrid load powers collected at the 1st, 2nd…nth collection moments respectively.

[0046] The state of microgrid adjustable resources can be expressed as shown in formula (6):

[0047]

[0048] In formula (6), Y FR is the time series vector of the adjustable resource state of the microgrid; They are the adjustable powers of the microgrid adjustable resources collected at the 1st, 2nd…nth collection moments respectively.

[0049] Therefore, the final microgrid emergency power control scenario set is

[0050]

[0051] Among them, Y WS Y is the microgrid emergency power control scenario set; DK is the microgrid control threshold.

[0052] Step 2: Construct a generative adversarial network model for microgrid regulation threshold setting. The generative adversarial network model includes an emergency power control scenario generator and a microgrid regulation threshold effect discriminator.

[0053] Gaussian noise is superimposed on the microgrid photovoltaic output data, microgrid wind power output data, microgrid load status, and microgrid adjustable resource status in the microgrid emergency power control scenario set obtained in step 1, thereby simulating the uncertainty of microgrid operation. Then, the microgrid emergency power control scenario set superimposed with Gaussian noise is input into the emergency power control scenario generator in the generative adversarial network model to perform multiple rounds of adversarial learning of the generative adversarial network model. The adversarial learning process is as follows:

[0054] (1) In each round of adversarial learning, the emergency power control scenario generator optimizes the parameters of its own reward function based on the improved plant cell swarm algorithm, and generates the current microgrid emergency power control scenario and the current microgrid regulation threshold corresponding to the current microgrid emergency power control scenario under the guidance of the optimized reward function.

[0055] In this embodiment, the emergency power control scenario generator is composed of multiple layers of fully connected layers and universal activation functions. When the microgrid emergency power control scenario set is input into the emergency power control scenario generator, each fully connected layer of the emergency power control scenario generator extracts the features of the microgrid emergency power control scenario set (including scenario status and threshold), and uses the universal activation function to determine whether the extracted features are transmitted to the next fully connected layer; if the judgment result is yes, the features extracted by the current fully connected layer can be transmitted to the next fully connected layer; the features are transmitted successively until the extracted microgrid emergency power control scenario set features are transmitted to the end of the emergency power control scenario generator, and the current emergency power control scenario set status and threshold are obtained. In addition, the current microgrid control threshold generated by the emergency power control scenario generator includes the control power shortage threshold and the control delay time.

[0056] In this embodiment, the reward function of the emergency power control scenario generator itself is shown in formula (7):

[0057] R(λ)=ω(λ n -λ n-1 )+b,2≤n≤N

[0058] Where: R(λ) is the reward function of the emergency power control scenario generator; n is the sample generated by the emergency power control scenario generator for the nth time; ω is the sensitivity coefficient of the emergency power control scenario generator. b is the bias coefficient of the emergency power control scenario generator; N is the total number of samples generated by the emergency power control scenario generator;

[0059] In each round of adversarial learning, the emergency power control scenario generator optimizes the sensitivity coefficient ω and the bias coefficient b in the reward function based on the improved plant cell swarm algorithm.

[0060] In this embodiment, the improved plant cell swarm algorithm is obtained by improving the growth position update on the basis of the existing plant cell swarm algorithm, and the global exploration capability and local mining capability of the plant cell swarm algorithm are dynamically adjusted when updating the growth position by adding a dynamic growth parameter in the growth position update function. The optimization solution process of the sensitivity coefficient ω and the bias coefficient b in the reward function of the emergency power control scenario generator by the improved plant cell swarm algorithm is as follows:

[0061] S1. Initialize the parameters of the improved plant cell swarm algorithm. The parameters of the improved plant cell swarm algorithm include: the number of plant cells of the improved plant cell swarm algorithm, the optimization dimension of the improved plant cell swarm algorithm, and the initial position of the plant cells of the improved plant cell swarm algorithm (initial random solution). Among them, the optimization dimension of the improved plant cell swarm algorithm is 2, including the sensitivity coefficient ω in the reward function of the emergency power control scenario generator and the bias coefficient b of the emergency power control scenario generator.

[0062] S2. Design and improve the growth position update function of plant cells in the plant cell swarm algorithm, and use the growth position update function to find the optimal solution of the sensitivity coefficient ω and the bias coefficient b of the emergency power control scenario generator reward function. The growth position update function is shown in formula (8):

[0063]

[0064] In formula (8): are the growth positions of plant cells at time t and t+1 in the improved plant cell swarm algorithm; d g To improve the random growth distance of plant cells in the plant cell swarm algorithm; G c To improve the dynamic growth constant in the plant cell swarm algorithm; To improve the optimal position of plant cells in the plant cell swarm algorithm, the growth position of plant cells at t.

[0065] S3. Use the updated plant cell growth position as the output solution of the sensitivity coefficient ω and the bias coefficient b of the emergency power control scenario generator reward function in the current iteration round.

[0066] (2) In each round of adversarial learning, after the emergency power control scenario generator generates the current microgrid regulation threshold, the microgrid regulation threshold effect discriminator determines whether the current microgrid regulation threshold meets the microgrid emergency power regulation requirements.

[0067] In this embodiment, the microgrid regulation threshold effect discriminator is composed of multiple layers of fully connected layers and a universal activation function. When the current emergency power control scenario set state and threshold generated by the emergency power control scenario generator are input into the microgrid regulation threshold effect discriminator, each fully connected layer of the microgrid regulation threshold effect discriminator extracts the current emergency power control scenario set state and the microgrid operation safety state under the threshold, and uses a universal activation function to determine whether the extracted features are transmitted to the next fully connected layer; if the judgment result is yes, the features extracted by the current fully connected layer can be transmitted to the next fully connected layer; the features are transmitted successively until the extracted microgrid operation safety state features are transmitted to the end of the microgrid regulation threshold effect discriminator to obtain the current microgrid operation safety state features, and then determine whether the current microgrid operation safety state features meet the microgrid operation requirements.

[0068] In this embodiment, if the judgment result of the microgrid control threshold effect discriminator is that the current microgrid control threshold cannot meet the microgrid emergency power control demand, the microgrid control threshold effect discriminator feeds back the judgment result to the emergency power control scenario generator, and the emergency power control scenario generator in the generative adversarial network model continues to optimize the parameters of its own reward function through the improved plant cell swarm algorithm, and generates a microgrid emergency power control scenario and a microgrid control threshold corresponding to the microgrid emergency power control scenario under the guidance of the optimized reward function, that is, repeats the above process to continue to iterate for the next round of adversarial learning.

[0069] If the judgment result is that the current microgrid control threshold can meet the microgrid emergency power control demand, the emergency power control scenario generator in the generative adversarial network model stops improving the plant cell swarm algorithm. At this time, the generative adversarial network model stops adversarial learning, and the emergency power control scenario generator outputs the current microgrid control threshold, thereby completing the microgrid control threshold setting.

[0070] In this embodiment, the optimization objective of the generative adversarial network model for microgrid control threshold setting is shown in formula (9):

[0071]

[0072] In formula (9), G represents the emergency power control scenario generator; D represents the microgrid regulation threshold effect discriminator; x is the data input to the emergency power control scenario generator; z is Gaussian noise; p z () is the normal distribution function; E() is the mathematical expectation function; V is the optimization objective function.

[0073] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present embodiment is further described in detail below in conjunction with experiments.

[0074] The operation data of a microgrid in a province in eastern my country throughout 2022 was selected for case analysis. Microgrid data Grid operation data and environmental meteorological data. The generative adversarial network is built based on the deep school framework, and the computer hardware for training includes: Intel I9 CPU, 32GB memory, GTX 4070TI GPU. The initial number of plant cells in the improved plant cell swarm algorithm is 100, the optimization dimension of the improved plant cell swarm algorithm is 2, and the initial position of the plant cells in the improved plant cell swarm algorithm is [0,0]. The maximum number of iterations of the improved plant cell swarm algorithm is set to 500.

[0075] In order to compare the performance of the method proposed in this embodiment, three different methods were set up for simulation analysis in the experimental case analysis. The results are as follows: Figure 2As shown. Method 1: the method proposed in this embodiment (proposed method curve in the figure); Method 2: the setting method based on the particle swarm algorithm (PSO algorithm) (PSO curve in the figure); Method 3: the setting method based on the genetic algorithm (GA algorithm) (GA curve in the figure). The horizontal axis in the figure represents the number of emergency power controls, and the vertical axis represents the completion effect of the emergency power control. It can be seen that the microgrid control threshold setting effect based on the method proposed in this embodiment is better, and can better enable the microgrid to respond to the emergency power control needs of the power grid.

[0076] In order to compare the optimization performance of the improved plant cell swarm algorithm proposed in this embodiment, two different methods were set up for simulation analysis in the experimental case analysis. The results are as follows: Figure 3 As shown. Method 1: the method proposed in this embodiment (the improved method curve in the figure); Method 2: the tuning method based on the traditional plant cell swarm algorithm (the traditional method curve in the figure). It can be seen that the improved plant cell swarm algorithm proposed in this embodiment can converge faster and better, thereby improving the overall optimization ability of the microgrid control threshold intelligent tuning algorithm.

[0077] The preferred embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. The embodiments described in the present invention are merely descriptions of the preferred embodiments of the present invention, and do not limit the concept and scope of the present invention. The various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction, and such combinations should also be regarded as the contents disclosed in the present disclosure as long as they do not violate the concept of the present invention. In order to avoid unnecessary repetition, the present invention will not further describe various possible combinations.

[0078] The present invention is not limited to the specific details of the above-mentioned embodiments. Within the technical concept of the present invention and without departing from the design concept of the present invention, various modifications and improvements made to the technical solution of the present invention by technical personnel in this field should fall within the protection scope of the present invention. The technical contents for which protection is sought in the present invention have been fully recorded in the claims.

Claims

1. A microgrid control threshold intelligent setting method for emergency power control, characterized in that: The following steps are involved: Step 1: Obtain the grid emergency power regulation demand information, microgrid output data, and microgrid status information, and construct a microgrid emergency power control scenario set based on the grid emergency power regulation demand information, microgrid output data, and microgrid status information; Step 2: construct a generative adversarial network model for microgrid regulation threshold setting, wherein the generative adversarial network model includes an emergency power control scenario generator and a microgrid regulation threshold effect discriminator, and the microgrid emergency power control scenario set obtained in step 1 is input into the emergency power control scenario generator in the generative adversarial network model to perform multi-round adversarial learning of the generative adversarial network model; In each round of adversarial learning, the emergency power control scenario generator optimizes and solves the parameters of its own reward function based on the improved plant cell swarm algorithm, and generates the current microgrid emergency power control scenario and the current microgrid regulation threshold corresponding to the current microgrid emergency power control scenario under the guidance of the optimized reward function, and the microgrid regulation threshold effect discriminator determines whether the current microgrid regulation threshold meets the microgrid emergency power regulation demand; If the judgment result of the microgrid regulation threshold effect discriminator is that the current microgrid regulation threshold cannot meet the microgrid emergency power regulation demand, the microgrid regulation threshold effect discriminator feeds back the judgment result to the emergency power control scenario generator, and generates an adversarial network model to repeat the above process for the next round of adversarial learning; If the judgment result is that the current microgrid control threshold can meet the microgrid emergency power control demand, the generative adversarial network model stops adversarial learning, and the emergency power control scenario generator outputs the current microgrid control threshold, thereby completing the microgrid control threshold setting.

2. According to claim 1, a microgrid control threshold intelligent setting method for emergency power control is characterized in that: In step 1, the microgrid output data includes microgrid photovoltaic output data and microgrid wind power output data.

3. The method for intelligently setting a microgrid control threshold for emergency power control according to claim 1 is characterized in that: In step 1, the microgrid status information includes the microgrid energy storage device status, the microgrid load status, and the microgrid adjustable resource status.

4. The method for intelligently setting a microgrid control threshold for emergency power control according to claim 1 is characterized in that: In step 2, after the noise is centrally superimposed on the microgrid emergency power control scenario, it is input into the emergency power control scenario generator to simulate the uncertainty of microgrid operation by superimposing the noise.

5. The method for intelligently setting a microgrid control threshold for emergency power control according to claim 1 is characterized in that: In step 2, the emergency power control scenario generator is composed of multiple layers of fully connected layers and a universal activation function. When the microgrid emergency power control scenario set is input into the emergency power control scenario generator, each fully connected layer of the emergency power control scenario generator extracts the features of the microgrid emergency power control scenario set, and uses the universal activation function to determine whether the extracted features are transmitted to the next fully connected layer; If the judgment result is yes, the features extracted by the current fully connected layer can be transferred to the next fully connected layer; The extracted microgrid emergency power control scenario set features are transmitted successively until they are transmitted to the end of the emergency power control scenario generator to obtain the current emergency power control scenario set state and threshold.

6. The method for intelligently setting a microgrid control threshold for emergency power control according to claim 1 is characterized in that: In step 2, the microgrid regulation threshold effect discriminator is composed of multiple layers of fully connected layers and a universal activation function. When the current emergency power control scenario set state and threshold generated by the emergency power control scenario generator are input into the microgrid regulation threshold effect discriminator, each fully connected layer of the microgrid regulation threshold effect discriminator extracts the current emergency power control scenario set state and the microgrid operation safety state under the threshold, and uses the universal activation function to determine whether the extracted features are transmitted to the next fully connected layer; If the judgment result is yes, the features extracted by the current fully connected layer can be transferred to the next fully connected layer; The extracted microgrid operation safety status characteristics are transmitted successively until they are transmitted to the end of the microgrid regulation threshold effect discriminator to obtain the current microgrid operation safety status characteristics, and then judge whether the current microgrid operation safety status characteristics meet the microgrid operation requirements.

7. The method for intelligently setting a microgrid control threshold for emergency power control according to claim 1 is characterized in that: In step 2, the current microgrid control threshold obtained by the emergency power control scenario generator includes a control power shortage threshold and a control delay time.

8. The method for intelligently setting a microgrid control threshold for emergency power control according to claim 1, characterized in that: In step 2, the reward function of the emergency power control scenario generator is as follows: R(λ)=ω(λ n -l n-1 )+b,2≤n≤N Where: R(λ) is the reward function of the emergency power control scenario generator; n is the sample generated by the emergency power control scenario generator for the nth time; ω is the sensitivity coefficient of the emergency power control scenario generator; b is the bias coefficient of the emergency power control scenario generator; N is the total number of samples generated by the emergency power control scenario generator; In each round of adversarial learning, the emergency power control scenario generator optimizes the sensitivity coefficient ω and the bias coefficient b in the reward function based on the improved plant cell swarm algorithm.

Images