CN118195160A

CN118195160A - Power grid section regulation and control model training method and device and computer equipment

Info

Publication number: CN118195160A
Application number: CN202410377961.8A
Authority: CN
Inventors: 陈兴望; 王坚; 邱生敏; 张坤; 吴小刚; 吕耀棠; 张艺镨; 李志中; 刘士齐; 何劲松
Original assignee: China Southern Power Grid Co Ltd
Current assignee: China Southern Power Grid Co Ltd
Priority date: 2024-03-29
Filing date: 2024-03-29
Publication date: 2024-06-14

Abstract

The application relates to a training method and device for a power grid section regulation model and computer equipment. The method comprises the following steps: acquiring training sample data of a section regulation model corresponding to a target power grid, and training the section regulation model of the power grid to be trained and a power grid training section optimization network; inputting the training sample data of the section control model into a power grid section control model to be trained to obtain initial power grid section control information corresponding to a target power grid; inputting the initial power grid section regulation and control information into a power grid section training optimization network to obtain optimized power grid section regulation and control information; and training the power grid section regulation model to be trained by using the optimized power grid section regulation information until the disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value, so as to obtain the trained power grid section regulation model. By adopting the method, more intelligent and flexible power grid management can be realized, so that the response speed and adaptability of the power system are improved, and the stability of the power system is further improved.

Description

Power grid section regulation and control model training method and device and computer equipment

Technical Field

The application relates to the technical field of smart power grids, in particular to a power grid section regulation and control model training method, a device, computer equipment, a storage medium and a computer program product.

Background

With the development of computer technology, smart grid technology appears, and the construction of a power grid section regulation model is an advanced technology applied to a power system, and can predict and optimize a regulation strategy of a power grid section based on real-time data and scene information of the power system so as to improve the stability, efficiency and reliability of the power system. The power grid regulation and control technology in the traditional technology mainly depends on experience rules and static models, is limited by model precision and instantaneity, and is difficult to adapt to complex power system changes. Resulting in poor stability of the power system.

Disclosure of Invention

In view of the foregoing, it is desirable to provide a power grid section control model training method, apparatus, computer device, computer readable storage medium, and computer program product that can improve the stability of a power system.

In a first aspect, the present application provides a method for training a power grid section control model, including:

Acquiring training sample data of a section regulation model corresponding to a target power grid, and training the section regulation model of the power grid to be trained and a power grid training section optimization network;

Inputting the training sample data of the section control model into the section control model of the power grid to be trained to obtain initial power grid section control information corresponding to the target power grid;

Inputting the initial power grid section regulation and control information into the power grid training section optimization network to obtain optimized power grid section regulation and control information;

and training the power grid section regulation model to be trained by using the optimized power grid section regulation information until the disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value, so as to obtain a trained power grid section regulation model.

In a second aspect, the present application further provides a power grid section control model training device, including:

the power grid data acquisition module is used for acquiring the training sample data of the section regulation model corresponding to the target power grid, the power grid section regulation model to be trained and the power grid section training optimization network;

The regulation and control information calculation module is used for inputting the training sample data of the section regulation and control model to the power grid section regulation and control model to be trained to obtain initial power grid section regulation and control information corresponding to the target power grid;

the regulation and control information optimizing module is used for inputting the initial power grid section regulation and control information into the power grid training section optimizing network to obtain optimized power grid section regulation and control information;

And the regulation and control model training module is used for training the power grid section regulation and control model to be trained by using the optimized power grid section regulation and control information until the disturbance of the initial power grid section regulation and control information is smaller than a disturbance threshold value, so as to obtain a trained power grid section regulation and control model.

In a third aspect, the present application also provides a computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:

In a fourth aspect, the present application also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of:

In a fifth aspect, the application also provides a computer program product comprising a computer program which, when executed by a processor, performs the steps of:

According to the method, the device, the computer equipment, the storage medium and the computer program product for training the power grid section regulation model, the power grid section regulation model to be trained and the power grid training section optimization network are obtained by obtaining the training sample data of the section regulation model corresponding to the target power grid; inputting the training sample data of the section control model into a power grid section control model to be trained to obtain initial power grid section control information corresponding to a target power grid; inputting the initial power grid section regulation and control information into a power grid section training optimization network to obtain optimized power grid section regulation and control information; and training the power grid section regulation model to be trained by using the optimized power grid section regulation information until the disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value, so as to obtain the trained power grid section regulation model.

Training a power grid section regulation model to be trained by utilizing the obtained section regulation model training sample data corresponding to the target power grid, obtaining initial power grid section regulation information, and inputting the initial power grid section regulation information into a power grid section optimization network to obtain optimized power grid section regulation information. And then, training the power grid section regulation model to be trained by utilizing the optimized power grid section regulation information until the disturbance of the power grid section regulation information is smaller than a set disturbance threshold value, thereby obtaining the trained power grid section regulation model. By training and optimizing the power grid section regulation and control model, the system can realize more intelligent and flexible power grid management, so that the response speed and adaptability of the power system are improved, various external and internal changes can be dealt with, and the stability, reliability and economy of the power system are further improved. And secondly, the intelligent section regulation and control can optimize the structure and power distribution of the power grid, reduce the energy loss and transmission loss, reduce the running cost of the system and improve the energy utilization efficiency. In addition, through continuous optimization and learning, the power grid section regulation and control model can continuously adapt to changing operation environments and market demands, an optimal regulation and control strategy is maintained, and important contribution is made to sustainable development and optimal operation of the power industry.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are required to be used in the embodiments or the related technical descriptions will be briefly described, and it is apparent that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for those skilled in the art.

FIG. 1 is an application environment diagram of a power grid section control model training method in one embodiment;

FIG. 2 is a flow chart of a method for training a grid section control model according to an embodiment;

FIG. 3 is a flow chart of a method for obtaining initial grid section control information in one embodiment;

FIG. 4 is a flow chart of a method for obtaining training power grid section control information in one embodiment;

FIG. 5 is a flow chart of a method for optimizing power grid section control information in one embodiment;

FIG. 6 is a flowchart of a method for obtaining optimal grid section control information according to another embodiment;

FIG. 7 is a flow chart of a method for obtaining a trained power grid section control model in one embodiment;

FIG. 8 is a block diagram of a power grid section control model training device in one embodiment;

fig. 9 is an internal structural diagram of a computer device in one embodiment.

Detailed Description

The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.

The power grid section regulation model training method provided by the embodiment of the application can be applied to an application environment shown in figure 1. Wherein the terminal 102 communicates with the server 104 via a network. The data storage system may store data that the server 104 needs to process. The data storage system may be integrated on the server 104 or may be located on a cloud or other network server. The server 104 obtains the training sample data of the section regulation model corresponding to the target power grid, the power grid section regulation model to be trained and the power grid section training optimization network through the terminal 102; inputting the training sample data of the section control model into a power grid section control model to be trained to obtain initial power grid section control information corresponding to a target power grid; inputting the initial power grid section regulation and control information into a power grid section training optimization network to obtain optimized power grid section regulation and control information; and training the power grid section regulation model to be trained by using the optimized power grid section regulation information until the disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value, so as to obtain the trained power grid section regulation model. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, internet of things devices, and portable wearable devices, where the internet of things devices may be smart speakers, smart televisions, smart air conditioners, smart vehicle devices, and the like. The portable wearable device may be a smart watch, smart bracelet, headset, or the like. The server 104 may be implemented as a stand-alone server or as a server cluster of multiple servers.

In an exemplary embodiment, as shown in fig. 2, a power grid section control model training method is provided, and an example of application of the method to the server in fig. 1 is described, which includes the following steps 202 to 208. Wherein:

step 202, obtaining training sample data of a section control model corresponding to a target power grid, a power grid section control model to be trained and a power grid section training optimization network.

The target power grid can be a power grid needing power grid section regulation and control.

The section control model training sample data can be training sample data for training a power grid section control model.

The power grid section control model to be trained can be a power grid section control model which is not trained yet.

The grid section training optimization network can be a model for optimizing output data of a grid section regulation model.

Specifically, historical operation data of a target power grid is obtained, the historical operation data comprise load data, generator data, line data and the like, a model of a power system is constructed according to the collected data, the model comprises a node model, a line model and the like, a proper section regulation scheme is designed, the power regulation range of each node, the power transmission capacity limit of the line and the like are included, and the historical data and the designed regulation scheme are utilized to generate section regulation model training sample data, wherein the section regulation model training sample data comprise input states and expected output control actions under different operation conditions. And designing a proper neural network structure as a model to be trained of a power grid section regulation model, selecting a deep neural network structure suitable for a power system regulation task, such as a multi-layer perceptron (MLP), a Convolutional Neural Network (CNN) or a cyclic neural network (RNN), initializing model parameters, and preparing for starting a training process. The structure of the optimization network is designed, and is usually an optimization algorithm for optimizing section regulation, such as a genetic algorithm, a particle swarm optimization algorithm, a reinforcement learning algorithm and the like, parameters of the optimization network are initialized, and a section optimization process is prepared to obtain the grid section training optimization network.

And 204, inputting the training sample data of the section control model into a power grid section control model to be trained, and obtaining initial power grid section control information corresponding to the target power grid.

The initial power grid section regulation information can be regulation information which is preliminarily calculated through a power grid section regulation model, but verification and modification of efficiency, safety and the like are not performed.

Specifically, the training sample data of the section control model is input to the section control model of the power grid to be trained in batches, so that the data input sequence is ensured to be consistent with the corresponding characteristics of the input layer of the model, and the data format is ensured to be consistent with the input requirement of the model. And carrying out forward propagation on training sample data of each section regulation model, and passing input data through each layer of the neural network to finally obtain initial power grid section regulation information corresponding to the target power grid. In the forward propagation process, the input data is multiplied by the weight and bias of the model, the output of each layer is calculated through an activation function, the predicted output of the model is compared with the actual label of the sample data, and the value of a loss function is calculated, wherein the commonly used loss function comprises Mean Square Error (MSE), cross entropy and the like and is used for measuring the difference between the predicted result of the model and the actual label.

And 206, inputting the initial power grid section regulation and control information into a power grid section training optimization network to obtain optimized power grid section regulation and control information.

The optimized power grid section regulation information can be initial power grid section regulation information after optimization.

Specifically, the initial power grid section regulation and control information is input into a power grid section training optimization network, and the power grid section training optimization network performs optimization calculation according to the input initial power grid section regulation and control information so as to improve the efficiency and performance of power grid operation. And calculating to obtain the optimized power grid section regulation and control information through an optimization algorithm and an objective function. Optimizing the power grid section regulation information may include node power distribution after adjustment, load distribution of a line, generator output adjustment and the like so as to realize balance and stability of the power grid.

And step 208, training the power grid section regulation model to be trained by using the optimized power grid section regulation information until the disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value, so as to obtain a trained power grid section regulation model.

The trained power grid section control model can be a trained power grid section control model.

Specifically, the power grid section regulation and control information is optimized, or the power grid section regulation and control model to be trained is trained by combining section regulation and control model training sample data, and the weight and bias of the model are updated through a back propagation algorithm so as to reduce the prediction error of the model. After each round of calculation, calculating the size between disturbance of the initial power grid section regulation information corresponding to the model and a disturbance threshold set by a target power grid, if the disturbance of the initial power grid section regulation information is smaller than the disturbance threshold, stopping training, and converging the model to a better state to obtain a trained power grid section regulation model; and otherwise, continuing the next round of training until the disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value to obtain a trained power grid section regulation model. The model in the training process is simultaneously evaluated, the performance and generalization capability of the model on the verification set are checked, and the prediction precision and accuracy of the model can be evaluated by comparing the model with actual operation data and using the verification set.

In the power grid section regulation model training method, training sample data of a section regulation model corresponding to a target power grid, a power grid section regulation model to be trained and a power grid section training optimization network are obtained; inputting the training sample data of the section control model into a power grid section control model to be trained to obtain initial power grid section control information corresponding to a target power grid; inputting the initial power grid section regulation and control information into a power grid section training optimization network to obtain optimized power grid section regulation and control information; and training the power grid section regulation model to be trained by using the optimized power grid section regulation information until the disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value, so as to obtain the trained power grid section regulation model.

In an exemplary embodiment, as shown in fig. 3, the section control model training sample data is input to the grid section control model to be trained to obtain initial grid section control information corresponding to the target grid, including steps 302 to 306. Wherein:

And 302, inputting the training sample data of the section control model into a power grid section control model to be trained to obtain training power grid section control information.

The training power grid section regulation information can be a calculation result of a power grid section regulation model, but the calculation result is not subjected to simulation verification and optimization.

Specifically, the training sample data of the section control model is input to the section control model of the power grid to be trained in batches, so that the data input sequence is ensured to be consistent with the corresponding characteristics of the input layer of the model, and the data format is ensured to be consistent with the input requirement of the model. The method comprises the steps of carrying out forward propagation on training sample data of each section regulation model, passing input data through each layer of a neural network, finally obtaining training power grid section regulation information corresponding to a target power grid, multiplying the input data by the weight and bias of the model in the forward propagation process, calculating the output of each layer through an activation function, comparing the predicted output of the model with the actual label of the sample data, calculating the value of a loss function, wherein the common loss function comprises Mean Square Error (MSE), cross entropy and the like, and is used for measuring the difference between a model predicted result and the actual label.

And step 304, importing the training power grid section regulation information into a power grid system simulation network to obtain power grid operation simulation data.

The grid system simulation network can be a model for checking whether the training grid section regulation information meets the standard of the target grid.

The power grid operation simulation data can be a result obtained after the training power grid section regulation information is checked.

Specifically, the section regulation information of the training power grid is input to a power grid system simulation network for simulation operation, and when the power grid system simulation network simulates the operation conditions of the power grid under different loads, power generation amounts, line states and the like according to the input regulation information, the randomness and the uncertainty in the simulation environment of the power grid system simulation network are considered so as to simulate the actual power grid operation conditions as truly as possible. In the simulation operation process of the power grid system, information such as various parameters, node states, line loads and the like of the simulation network is recorded, and finally power grid operation simulation data are output.

And 306, training the power grid section regulation model to be trained according to the power grid operation simulation data until the power grid operation simulation data accords with the power operation standard, and taking the training power grid section regulation information as initial power grid section regulation information.

The power operation standard may be an operation standard specified by the target grid.

Specifically, the power grid operation simulation data are input into a power grid section regulation model to be trained, and parameters of the model are continuously adjusted through a back propagation algorithm, so that the model gradually approaches to a mode and a standard of actual power grid operation. After each round of calculation, the performance of the model is evaluated by using a verification set or a cross verification method and the like, indexes such as accuracy, loss value and the like of the model between power grid operation simulation data and power operation standards are checked, and super-parameters and structures of the model are adjusted so as to improve generalization capability and prediction accuracy of the model. If the power grid operation simulation data accords with the power operation standard, stopping training, taking the training power grid section regulation information just obtained as initial power grid section regulation information, otherwise, continuously repeating training until the power grid operation simulation data accords with the power operation standard, and taking the training power grid section regulation information as initial power grid section regulation information. The training process can set the training round number, the learning rate attenuation strategy and the like according to the needs so as to control the convergence speed and the stability of the training process.

In this embodiment, training sample data of the section control model is input into the power grid section control model to be trained to obtain training power grid section control information, and then the information is imported into a power grid system simulation network to obtain power grid operation simulation data. And training the power grid section regulation model to be trained by using the simulation data until the simulation data accords with the electric power operation standard. The method has the main advantages that the power grid regulation strategy can be effectively optimized through training of the simulation data, and the running efficiency and stability of the power grid are improved. Meanwhile, the model parameters are continuously adjusted, so that the model can be better adapted to different power running conditions, and more reliable support and guidance are provided for actual power grid running.

In one exemplary embodiment, as shown in FIG. 4, the grid section regulation model to be trained includes a grid regulation strategy network and a grid regulation value network; the training sample data of the section control model is input into the section control model of the power grid to be trained to obtain the section control information of the training power grid, and the training power grid comprises steps 402 to 410. Wherein:

Step 402, inputting the section regulation model training sample data into a power grid regulation strategy network to obtain power grid regulation decision information.

The grid regulation strategy network may be a model for generating a regulation strategy from data samples.

The power grid regulation decision information can be strategy information for regulating and controlling a power grid section regulation model to be trained.

Specifically, the section control model training sample data is input into a power grid control strategy network, the dimension and format of the input data are required to be matched with the input layer of the network, the section control model training sample data are transmitted from the input layer to the output layer through a forward propagation algorithm, and the output result of the network is calculated. After each node receives the input, the result is transmitted to the next layer of nodes through the activation function processing until the next layer of nodes are output, and the regulation and control decision information of the power grid is calculated according to the output result of the network.

And step 404, inputting the training sample data of the section regulation model into a power grid regulation value network to obtain a comprehensive rewarding value of the power grid.

Wherein the grid regulation value network may be a model for calculating regulated operation costs from data samples

The comprehensive power grid rewarding value can be running cost information regulated and controlled by a power grid section regulation model to be trained.

Specifically, the section control model training sample data is input into a power grid control value network, the dimension and format of the input data are required to be matched with the input layer of the network, the section control model training sample data are transmitted from the input layer to the output layer through a forward propagation algorithm, and the output result of the network is calculated. After each node receives the input, the result is transmitted to the next layer of nodes through the activation function processing until reaching the output layer, and the comprehensive rewarding value of the power grid is calculated according to the output result of the network.

And step 406, adjusting the power grid regulation and control decision information and the power grid comprehensive rewards value by using the power grid stability simulation model to obtain the regulation and control decision information and the comprehensive rewards value.

The grid stability simulation model may be used to simulate the stability of the input parameters applied on the target point grid.

The regulation decision information and the comprehensive rewards value can be respectively obtained regulation values for the power grid regulation decision information and the comprehensive rewards value of the power grid.

Specifically, a grid stability simulation model is invoked, which may be a physical simulation-based model or a machine learning-based model, for assessing the stability of the grid. And training or parameter tuning is carried out on the power grid stability simulation model, so that the power grid stability simulation model can accurately simulate the running condition of the power grid. And simulating the power grid regulation and control decision information and the power grid comprehensive rewarding value by using a power grid stability simulation model, and analyzing the stability performance of the target power grid under different decision schemes. And evaluating the feasibility and rationality of each decision scheme and the influence on the stability of the power grid according to the actual running condition and constraint conditions of the power grid. And generating regulation and control decision information and a comprehensive rewarding value according to the simulation result.

And step 408, adjusting model parameters of the power grid regulation strategy network and the power grid regulation value network according to the regulation decision information and the regulation comprehensive rewards value to obtain a parameter regulation power grid section regulation model.

The parameter adjustment power grid section control model can be a power grid section control model to be trained which is subjected to parameter adjustment but fails to meet the use requirement.

Specifically, according to the regulation and control decision information and the regulation comprehensive rewarding value, the model parameters of the power grid regulation and control strategy network and the power grid regulation and control value network are respectively regulated, which can be realized through optimization algorithms such as gradient descent algorithm, genetic algorithm and the like, and a parameter regulation and control power grid section regulation model is obtained; and simultaneously calculating the gradient of the loss function or the objective function according to the adjustment comprehensive rewarding value, and updating the parameters of the network according to the gradient descending direction.

Step 410, the section control model training sample data is input to the parameter adjustment power grid section control model, and training power grid section control information is obtained.

Specifically, the training sample data of the section control model is input into the section control model of the parameter adjustment power grid, the input data is transmitted from an input layer to an output layer through a forward propagation algorithm, and the output result of the network is calculated. After each node receives the input, the result is transmitted to the next layer of nodes through the activation function processing until the next layer of nodes is output. And calculating to obtain the section regulation information of the training power grid according to the output result of the network.

In this embodiment, training sample data of the section regulation model is input into a power grid regulation strategy network and a power grid regulation value network to obtain power grid regulation decision information and a power grid comprehensive rewarding value, the decision information and the rewarding value are regulated by using a power grid stability simulation model to obtain regulated decision information and rewarding value, and model parameters of the power grid regulation strategy network and the power grid regulation value network are regulated according to the regulated information, so that a power grid section regulation model with regulated parameters is obtained, and training sample data is input into the power grid section regulation model with regulated parameters to obtain training power grid section regulation information. The method has the main advantages that the accuracy and the adaptability of the power grid regulation strategy are improved by dynamically adjusting the decision information and the rewarding value and optimizing the model parameters, so that the stability and the efficiency of the power grid are enhanced. Meanwhile, the model parameters are continuously optimized, so that the model can better adapt to the change and the demand of the power grid, and powerful support and guarantee are provided for the long-term stable operation of the power grid.

In an exemplary embodiment, as shown in fig. 5, the initial grid section regulation information is input to a grid training section optimization network to obtain optimized grid section regulation information, which includes steps 502 to 506. Wherein:

Step 502, randomly selecting a plurality of electric power safety consideration factors from a power grid system simulation network according to the initial power grid section regulation and control information.

The power safety consideration may be a safety factor that needs to be considered by the target power grid.

Specifically, according to the requirements and targets of grid regulation, power safety factors to be considered are determined, and the factors may include node voltage stability, line flow limit, generator output limit and the like. For each selected electric power safety consideration factor, according to the initial power grid section regulation and control information, randomly selecting corresponding nodes, lines or generators and the like from a power grid system simulation network, wherein the selection process can be realized through a random number generator, and randomness and representativeness are ensured. For each selected safety consideration, a particular parameter or indicator to be considered is determined. These parameters may include voltage values of the nodes, tidal current values of the lines, output values of the generator, etc.

Step 504, according to the initial power grid section regulation information, randomly generating a plurality of power efficiency calculation functions from a power grid system simulation network.

Wherein the power efficiency calculation function may be a function for calculating the efficiency of the target grid.

Specifically, according to the target and the requirement of the grid regulation, the type and the kind of the power efficiency calculation function to be generated are determined, and the calculation functions can comprise a node voltage stability index, a line power flow loss index, a generator efficiency index and the like. And randomly selecting a corresponding function form from a predefined function library according to the initial power grid section regulation information aiming at each type of power efficiency index. For example, a polynomial function, an exponential function, a logarithmic function, or the like may be selected as the form of the power efficiency calculation function. For a selected computational function, parameters in the function are determined, including coefficients, exponents, constant terms, and the like. These parameters can be generated by a random number generator within a certain range to ensure the diversity and representativeness of the function. And generating a corresponding power efficiency calculation function according to the selected function form and parameters.

Step 506, optimizing the initial power grid section regulation information according to the power safety consideration factors and the power efficiency calculation functions to obtain optimized power grid section regulation information.

Specifically, according to the target and the requirement of the power grid regulation, the target and the index to be optimized are determined, and the indexes can comprise safety indexes, efficiency indexes or the comprehensive consideration of the safety indexes and the efficiency indexes. For each selected power safety consideration factor and each power efficiency calculation function, corresponding safety and efficiency index values are calculated by combining the initial power grid section regulation information, and the index values are used for evaluating the quality of the current power grid section regulation information. According to the obtained safety and efficiency index values, the importance and the priority of each index are comprehensively considered to optimize the initial power grid section regulation and control information, and a heuristic algorithm, a genetic algorithm, a simulated annealing algorithm and the like can be adopted for the optimization method, so that the objective function is maximized or minimized. And obtaining the optimized power grid section regulation and control information after processing by an optimization algorithm, wherein the optimized power grid section regulation and control information comprises a node regulation scheme, the power transmission capacity of a line and the like.

In this embodiment, by randomly selecting power safety consideration factors from a power grid system simulation network and generating a power efficiency calculation function, the safety and efficiency of the power grid are comprehensively considered, and the initial power grid section regulation information is optimized according to the safety consideration factors and the efficiency calculation function, so as to obtain optimized power grid section regulation information. The method has the main advantages that the regulation strategy of the power grid is optimized by comprehensively considering the safety and efficiency factors, and the safety and the operation efficiency of the power grid are improved. Meanwhile, the optimization process is more diversified and comprehensive through the random selection and generation modes, the complexity and uncertainty of the power grid can be better dealt with, and more reliable guarantee is provided for the stable operation of the power grid.

In an exemplary embodiment, as shown in fig. 6, the initial grid section regulation information is optimized according to each power safety consideration factor and each power efficiency calculation function, so as to obtain optimized grid section regulation information, which includes steps 602 to 606. Wherein:

step 602, calculating load fluctuation information and new energy fluctuation information of the target power grid according to the power safety consideration factors and the initial power grid section regulation information.

The load fluctuation information can be the fluctuation condition of the load of the non-new energy generator set of the target power grid.

The new energy fluctuation information can be the fluctuation condition of the load of the new energy generator set of the target power grid.

Specifically, according to the electric power safety consideration factors and the initial power grid section regulation information, load fluctuation information of the target power grid is calculated by combining a load prediction model or historical load data, wherein the load fluctuation information may comprise a fluctuation range, frequency, time period and the like of a load curve. And calculating new energy fluctuation information of the target power grid according to the power safety consideration factors and the initial power grid section regulation and control information and combining a new energy power generation model or historical new energy power generation data, wherein the new energy fluctuation information can comprise fluctuation range, frequency, time period and the like of new energy power generation amounts of wind power, photovoltaic and the like. The load fluctuation information and the new energy fluctuation information are comprehensively considered and analyzed to evaluate the stability and reliability of the power grid, and simulation software can be used for simulation analysis or actual data can be adopted for statistical analysis in consideration of the dynamic characteristics and complexity of the power system.

Step 604, calculating efficiency balance fluctuation information of the target power grid according to the power efficiency calculation function, the initial power grid section regulation information, the load fluctuation information and the new energy fluctuation information.

The efficiency balance fluctuation information can be fluctuation conditions when the loads of all generator sets of the target power grid are balanced.

Specifically, according to the electric power efficiency calculation function, the initial power grid section regulation information, the load fluctuation information and the new energy fluctuation information, the efficiency and the balance of the power grid are comprehensively considered. According to the index value of the calculation function, the influence of the initial power grid section regulation information, the load fluctuation and the new energy fluctuation is combined, and the efficiency balance condition of the power grid is evaluated and analyzed to obtain the efficiency balance fluctuation information. Comprehensively analyzing the calculated efficiency balance fluctuation information, evaluating the running condition of the power grid, and if the condition of large or unstable efficiency balance fluctuation exists, considering the regulation strategy or optimization measures for adjusting the power grid so as to improve the efficiency and balance of the power grid.

And step 606, optimizing the initial power grid section regulation information according to the load fluctuation information, the new energy fluctuation information and the efficiency balance fluctuation information to obtain optimized power grid section regulation information.

Specifically, the influence of load fluctuation information, new energy fluctuation information and efficiency balance fluctuation information on the operation of the power grid is comprehensively considered, the aspects of stability, efficiency, balance and the like of the power grid are included, and the initial power grid section regulation and control information is optimized and adjusted, so that the power grid can better adapt to load fluctuation, new energy fluctuation and maintain efficiency balance. The optimization may involve aspects of node adjustment schemes, transmission capacity adjustment of lines, operation strategies of energy storage devices and the like, and the optimized power grid section regulation information is generated according to the adjusted power grid section regulation information.

In this embodiment, load balance and energy utilization efficiency of the power grid are comprehensively considered by calculating load fluctuation information and new energy fluctuation information of the target power grid and efficiency balance fluctuation information, and initial power grid section regulation information is optimized according to the fluctuation information, so that optimized power grid section regulation information is obtained. The method has the main advantages that the section regulation strategy of the power grid is optimized by comprehensively considering factors such as load fluctuation, new energy fluctuation, efficiency balance fluctuation and the like, so that the power grid can run more stably and efficiently in the face of different loads and energy fluctuation. Meanwhile, through calculation and analysis of fluctuation information, dynamic change of power grid operation can be better grasped, more accurate and reliable guidance is provided for power grid regulation and control, and overall performance and operation quality of the power grid are improved.

In an exemplary embodiment, as shown in fig. 7, the grid section regulation model to be trained is trained using optimized grid section regulation information until the disturbance of the initial grid section regulation information is less than a disturbance threshold, resulting in a trained grid section regulation model, comprising steps 702 to 704. Wherein:

Step 702, under the condition that the disturbance of the initial power grid section regulation information is greater than the disturbance threshold, according to the optimized power grid section regulation information, the power grid scene parameters, the regulation safety parameters, the regulation efficiency parameters and the scene generalization parameters of the power grid section regulation model to be trained are adjusted, and the adjusted power grid section regulation model is obtained.

Specifically, when the disturbance of the initial power grid section regulation information is greater than the disturbance threshold, according to the optimized power grid section regulation information, power grid scene parameters in a power grid section regulation model to be trained, including a topological structure, load distribution, new energy distribution and the like of a power grid, may need to be updated, and parameters related to power grid scenes, such as connection relations among nodes, parameters of a line, capacity of a generator and the like, may need to be updated. According to the optimized power grid section regulation and control information, regulating and control safety parameters in a power grid section regulation and control model to be trained, including rated power, power flow limit, voltage limit and the like of equipment, and possibly updating parameters related to regulation and control safety, such as rated power of equipment, power transmission capacity limit of a line, voltage limit of a node and the like. According to the optimized power grid section regulation and control information, regulating and control efficiency parameters in a power grid section regulation and control model to be trained, including the regulation speed, response time and the like of equipment, and parameters related to the regulation and control efficiency, such as the start-stop response time of a generator, the charge-discharge rate of energy storage equipment and the like, may need to be updated. According to the optimized power grid section regulation and control information, scene generalization parameters in a power grid section regulation and control model to be trained are adjusted so as to adapt to different power grid scenes, and parameters for generalization to different scenes, such as learning rate, regularization parameters and the like in the model, may need to be updated. And obtaining an adjusted power grid section regulation model through adjusting power grid scene parameters, regulation safety parameters, regulation efficiency parameters and scene generalization parameters.

And step 704, taking the adjusted power grid section regulation model as a power grid section regulation model to be trained, and returning to execute the step of inputting training sample data of the section regulation model into the power grid section regulation model to be trained to obtain initial power grid section regulation information corresponding to the target power grid until disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value to obtain the trained power grid section regulation model.

Specifically, the adjusted power grid section regulation model is used as a power grid section regulation model to be trained, and the step of inputting the training sample data of the section regulation model to the power grid section regulation model to be trained to obtain initial power grid section regulation information corresponding to a target power grid is performed in a return mode, if disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value, the trained power grid section regulation model is obtained, if disturbance of the initial power grid section regulation information is still larger than the disturbance threshold value, training is continued until the disturbance of the initial power grid section regulation information is smaller than the disturbance threshold value, and the trained power grid section regulation model is obtained.

In this embodiment, parameters of the power grid section regulation model are dynamically adjusted to adapt to changes in the running state of the power grid. When the disturbance of the initial power grid section regulation information is larger than a set disturbance threshold, the system adjusts the power grid scene parameters, the regulation safety parameters, the regulation efficiency parameters and the scene generalization parameters of the power grid section regulation model to be trained according to the optimized power grid section regulation information to obtain an adjusted power grid section regulation model. The optimization adjustment can enable the power grid section regulation and control model to be more suitable for the current running state of the power grid, and the flexibility and the robustness of power grid regulation and control are improved. And then, taking the adjusted power grid section regulation and control model as a model to be trained, and re-training the model until disturbance of the initial power grid section regulation and control information is smaller than a set disturbance threshold value, thereby obtaining a trained power grid section regulation and control model. The process ensures the high adaptation of the power grid section regulation model and the actual power grid running state, and effectively improves the stability and efficiency of the power grid.

It should be understood that, although the steps in the flowcharts related to the above embodiments are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.

Based on the same inventive concept, the embodiment of the application also provides a power grid section regulation model training device for realizing the power grid section regulation model training method. The implementation scheme of the device for solving the problem is similar to that described in the above method, so the specific limitation in the embodiments of the device for training the power grid section control model provided below may be referred to the limitation of the method for training the power grid section control model hereinabove, and will not be described herein.

In an exemplary embodiment, as shown in fig. 8, there is provided a power grid section control model training apparatus, including: a grid data acquisition module 802, a regulation information calculation module 804, a regulation information optimization module 806, and a regulation model training module 808, wherein:

the power grid data acquisition module 802 is configured to acquire training sample data of a section control model corresponding to a target power grid, a power grid section control model to be trained, and a power grid training section optimization network;

The regulation and control information calculation module 804 is configured to input the training sample data of the section regulation and control model to the power grid section regulation and control model to be trained, so as to obtain initial power grid section regulation and control information corresponding to the target power grid;

The regulation and control information optimizing module 806 is configured to input initial power grid section regulation and control information to a power grid training section optimizing network to obtain optimized power grid section regulation and control information;

And the regulation model training module 808 is configured to train the power grid section regulation model to be trained by using the optimized power grid section regulation information until the disturbance of the initial power grid section regulation information is less than the disturbance threshold value, thereby obtaining a trained power grid section regulation model.

In one embodiment, the regulation information calculation module 804 is further configured to input the section regulation model training sample data to a power grid section regulation model to be trained, so as to obtain training power grid section regulation information; importing the training power grid section regulation information into a power grid system simulation network to obtain power grid operation simulation data; training the power grid section regulation model to be trained according to the power grid operation simulation data until the power grid operation simulation data accords with the power operation standard, and taking the training power grid section regulation information as initial power grid section regulation information.

In one embodiment, the regulation information calculation module 804 is further configured to input the section regulation model training sample data to a power grid regulation policy network to obtain power grid regulation decision information; inputting the training sample data of the section regulation model into a power grid regulation value network to obtain a comprehensive rewarding value of the power grid; adjusting the power grid regulation and control decision information and the power grid comprehensive rewards value by using a power grid stability simulation model to obtain adjustment regulation and control decision information and adjustment comprehensive rewards value; adjusting model parameters of the power grid regulation strategy network and the power grid regulation value network according to the regulation and control decision information and the regulation comprehensive rewarding value to obtain a parameter regulation power grid section regulation model; and inputting the section control model training sample data into the parameter adjustment power grid section control model to obtain training power grid section control information.

In one embodiment, the regulation information optimization module 806 is further configured to randomly select a plurality of power safety consideration factors from the grid system simulation network according to the initial grid section regulation information; according to the initial power grid section regulation and control information, randomly generating a plurality of power efficiency calculation functions from a power grid system simulation network; and optimizing the initial power grid section regulation information according to the power safety consideration factors and the power efficiency calculation functions to obtain optimized power grid section regulation information.

In one embodiment, the regulation information optimization module 806 is further configured to calculate load fluctuation information and new energy fluctuation information of the target power grid according to the power safety consideration factor and the initial power grid section regulation information; calculating efficiency balance fluctuation information of a target power grid according to the power efficiency calculation function, the initial power grid section regulation information, the load fluctuation information and the new energy fluctuation information; and optimizing the initial power grid section regulation information according to the load fluctuation information, the new energy fluctuation information and the efficiency balance fluctuation information to obtain optimized power grid section regulation information.

In one embodiment, the regulation model training module 808 is further configured to, when the disturbance of the initial power grid section regulation information is greater than the disturbance threshold, adjust, according to the optimized power grid section regulation information, a power grid scene parameter, a regulation safety parameter, a regulation efficiency parameter, and a scene generalization parameter of the power grid section regulation model to be trained, to obtain an adjusted power grid section regulation model; and taking the adjusted power grid section regulation model as a power grid section regulation model to be trained, and returning to execute the step of inputting training sample data of the section regulation model into the power grid section regulation model to be trained to obtain initial power grid section regulation information corresponding to a target power grid until disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value to obtain the trained power grid section regulation model.

All or part of each module in the power grid section regulation model training device can be realized by software, hardware and a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.

In one exemplary embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 9. The computer device includes a processor, a memory, an Input/Output interface (I/O) and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is for storing server data. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for communicating with an external terminal through a network connection. The computer program when executed by the processor is used for realizing a power grid section regulation model training method.

It will be appreciated by persons skilled in the art that the architecture shown in fig. 9 is merely a block diagram of some of the architecture relevant to the present inventive arrangements and is not limiting as to the computer device to which the present inventive arrangements are applicable, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.

In an embodiment, there is also provided a computer device comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the method embodiments described above when the computer program is executed.

In one embodiment, a computer-readable storage medium is provided, storing a computer program which, when executed by a processor, implements the steps of the method embodiments described above.

In one embodiment, a computer program product or computer program is provided that includes computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the steps in the above-described method embodiments.

It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and the data (including but not limited to data for analysis, stored data, presented data, etc.) related to the present application are both information and data authorized by the user or sufficiently authorized by each party, and the collection, use and processing of the related data are required to meet the related regulations.

Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magneto-resistive random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (PHASE CHANGE Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in various forms such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The databases referred to in the embodiments provided herein may include at least one of a relational database and a non-relational database. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processor referred to in the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, or the like, but is not limited thereto.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The foregoing examples illustrate only a few embodiments of the application and are described in detail herein without thereby limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of the application should be assessed as that of the appended claims.

Claims

1. The power grid section regulation model training method is characterized by comprising the following steps of:

2. The method according to claim 1, wherein the inputting the section control model training sample data to the grid section control model to be trained to obtain the initial grid section control information corresponding to the target grid includes:

inputting the training sample data of the section control model to the section control model of the power grid to be trained to obtain training power grid section control information;

importing the training power grid section regulation information into a power grid system simulation network to obtain power grid operation simulation data;

And training the power grid section regulation model to be trained according to the power grid operation simulation data until the power grid operation simulation data accords with an electric power operation standard, and taking the training power grid section regulation information as the initial power grid section regulation information.

3. The method according to claim 2, wherein the grid section regulation model to be trained comprises a grid regulation strategy network and a grid regulation value network; the training sample data of the section control model is input to the section control model of the power grid to be trained to obtain training power grid section control information, and the training power grid section control information comprises:

Inputting the training sample data of the section regulation model into the power grid regulation strategy network to obtain power grid regulation decision information;

Inputting the training sample data of the section regulation model into the power grid regulation value network to obtain a comprehensive rewarding value of the power grid;

Using a power grid stability simulation model to adjust the power grid regulation and control decision information and the power grid comprehensive rewards value to obtain adjustment regulation and control decision information and adjustment comprehensive rewards value;

Adjusting model parameters of the power grid regulation strategy network and the power grid regulation value network according to the regulation decision information and the regulation comprehensive rewards value to obtain a parameter regulation power grid section regulation model;

and inputting the training sample data of the section control model to the parameter adjustment power grid section control model to obtain the training power grid section control information.

4. The method according to claim 1, wherein the inputting the initial grid section control information into the grid training section optimization network to obtain optimized grid section control information includes:

According to the initial power grid section regulation information, randomly selecting a plurality of power safety consideration factors from a power grid system simulation network;

according to the initial power grid section regulation information, randomly generating a plurality of power efficiency calculation functions from a power grid system simulation network;

and optimizing the initial power grid section regulation information according to the power safety consideration factors and the power efficiency calculation functions to obtain the optimized power grid section regulation information.

5. The method of claim 4, wherein optimizing the initial grid section control information according to each of the power safety considerations and each of the power efficiency calculation functions to obtain the optimized grid section control information comprises:

Calculating load fluctuation information and new energy fluctuation information of the target power grid according to the power safety consideration factors and the initial power grid section regulation information;

Calculating efficiency balance fluctuation information of the target power grid according to the power efficiency calculation function, the initial power grid section regulation information, the load fluctuation information and the new energy fluctuation information;

And optimizing the initial power grid section regulation information according to the load fluctuation information, the new energy fluctuation information and the efficiency balance fluctuation information to obtain the optimized power grid section regulation information.

6. The method according to claim 1, wherein training the grid section regulation model to be trained using the optimized grid section regulation information until the disturbance of the initial grid section regulation information is less than a disturbance threshold value, to obtain a trained grid section regulation model, comprises:

When the disturbance of the initial power grid section regulation information is larger than a disturbance threshold value, according to the optimized power grid section regulation information, adjusting power grid scene parameters, regulation safety parameters, regulation efficiency parameters and scene generalization parameters of the power grid section regulation model to be trained to obtain an adjusted power grid section regulation model;

And taking the adjusted power grid section regulation model as the power grid section regulation model to be trained, and returning to execute the step of inputting the training sample data of the section regulation model to the power grid section regulation model to be trained to obtain initial power grid section regulation information corresponding to the target power grid until disturbance of the initial power grid section regulation information is smaller than a disturbance threshold value to obtain the trained power grid section regulation model.

7. A power grid section regulation model training device, the device comprising:

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.

9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.