Virtual power plant control system and method

Abstract

The invention relates to the technical field of power dispatching, in particular to a virtual power plant control system and method. Comprising a data fusion module, a dynamic optimization scheduling module, an intelligent prediction module, a user interaction module, a self-learning adaptive module, a cooperative control module, a safety protection module and a storage unit. Key information is accurately extracted; the dynamic optimization scheduling module is based on a deep reinforcement learning algorithm, fully considers various factors to generate a global optimal strategy, and overcomes the defects of local optimization; the safety protection module constructs a multi-layer protection system to guarantee the safety of the system; the user interaction module covers the functions of information display, demand response, service customization and the like, personalized demands of users are met, complex conditions of the energy market and power grid operation can be handled on the whole, the operation benefit and safety of the virtual power plant are improved, and popularization is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of power dispatching, and in particular to a virtual power plant control system and method. Background Art

[0002] With the advancement of energy transformation, distributed energy resources have developed rapidly. Virtual power plants have emerged as a technical concept for effectively integrating multiple distributed energy sources. The existing virtual power plant control system has achieved centralized management and coordinated control of distributed energy to a certain extent. It can aggregate decentralized solar power stations, wind farms, small hydropower stations and various energy storage equipment, and realize information interaction and collaborative operation through advanced communication technology and software platforms, thereby improving energy utilization efficiency, reducing dependence on traditional fossil energy, providing strong support for the stable operation of the power grid, and helping to achieve sustainable energy supply and green development.

[0003] However, in the existing technology, firstly, in terms of data processing, faced with massive, multi-source, and heterogeneous energy data, the existing data processing methods are difficult to extract key information efficiently and accurately, resulting in decision-making delays and insufficient accuracy, and unable to respond to the complex and changing energy market and power grid operation needs in a timely manner; secondly, in terms of optimization and scheduling strategies, most systems are only based on simple rules or local optimization algorithms, and fail to fully consider the dynamic characteristics of energy equipment, uncertainty factors, and the coupling relationship between different energy forms, so that the operating efficiency of the entire virtual power plant fails to reach the optimal level; thirdly, the safe and reliable operation guarantee mechanism for virtual power plants is still imperfect, lacking a comprehensive security protection system and real-time risk warning and response measures, and is easily affected by abnormal situations such as network attacks and equipment failures, which in turn threaten the stable operation of the power grid; in addition, the existing system is relatively weak in terms of interaction with users, and cannot well meet the personalized energy service needs of users, which limits the widespread application and market promotion of virtual power plants. Summary of the Invention

[0004] The purpose of the present invention is to provide a virtual power plant control system and method, aiming to solve the problems in the existing technology. First, in terms of data processing, faced with massive, multi-source, and heterogeneous energy data, the existing data processing methods are difficult to extract key information efficiently and accurately, resulting in decision-making delays and insufficient accuracy, and unable to respond to the complex and changing energy market and power grid operation needs in a timely manner; secondly, in terms of optimization and scheduling strategies, most systems are only based on simple rules or local optimization algorithms, and fail to fully consider the dynamic characteristics of energy equipment, uncertainty factors, and the coupling relationship between different energy forms, so that the operating efficiency of the entire virtual power plant cannot reach the optimal level; thirdly, the safe and reliable operation guarantee mechanism for virtual power plants is still imperfect, lacks a comprehensive safety protection system and real-time risk warning and response measures, and is easily affected by abnormal situations such as network attacks and equipment failures, thereby threatening the technical problems of the power grid.

[0005] To achieve the above-mentioned purpose, the present invention adopts a virtual power plant control system, comprising a data fusion module, a dynamic optimization scheduling module, an intelligent prediction module and a user interaction module, wherein the data fusion module is connected to the dynamic optimization scheduling module, the dynamic optimization scheduling module is connected to the intelligent prediction module, the intelligent prediction module is connected to the user interaction module, the dynamic optimization scheduling module and the intelligent prediction module are connected to a self-learning adaptive module, the dynamic optimization scheduling module is connected to a collaborative control module, and the data fusion module, the dynamic optimization scheduling module, the intelligent prediction module and the user interaction module are connected to a security protection module and a storage unit;

[0006] The data fusion module is used to collect and integrate multi-source heterogeneous data from different types of distributed energy terminals, power grid monitoring equipment, and external energy markets, and build a unified energy data model through data cleaning, feature extraction, and fusion processing;

[0007] The dynamic optimization scheduling module is used to generate a global optimal energy scheduling strategy based on the energy data model, combined with real-time electricity price signals, grid load demand, and the operating status and constraints of each distributed energy device using a deep reinforcement learning algorithm;

[0008] The intelligent prediction module is used to use deep learning neural networks to make high-precision predictions of distributed energy power output, load change trends, and market price fluctuations;

[0009] The user interaction module is used to provide a visual user interface;

[0010] The self-learning and adaptive module is used to automatically adjust the parameters of the optimization scheduling model and the prediction model according to the data feedback and performance evaluation results during the actual operation of the virtual power plant;

[0011] The collaborative control module is used to decompose the strategy of the dynamic optimization scheduling module into specific control instructions and send them to each distributed energy device through a reliable communication protocol;

[0012] The security protection module is used to build a multi-layer security protection system;

[0013] The storage unit is used to store implementation data.

[0014] The data fusion module includes a data acquisition subunit, a data preprocessing subunit and a feature engineering subunit, and the data acquisition subunit, the data preprocessing subunit and the feature engineering subunit are all connected to the dynamic optimization scheduling module;

[0015] The data acquisition subunit is responsible for collecting energy data from various sensors, smart meters and third-party data platforms;

[0016] The data preprocessing subunit is used to perform denoising, normalization, and missing value filling on the collected data;

[0017] The feature engineering subunit is used to extract key features from the data and perform dimensionality reduction.

[0018] The dynamic optimization scheduling module includes an environment perception subunit, a strategy generation subunit and a strategy evaluation subunit, and the environment perception subunit, the strategy generation subunit and the strategy evaluation subunit are all connected to the intelligent prediction module;

[0019] The environmental sensing subunit is used to monitor the power grid operation status, energy market prices and policy information in real time;

[0020] The strategy generation subunit is used to generate scheduling strategies for different time scales and scenarios based on the model constructed by the deep reinforcement learning algorithm;

[0021] The strategy evaluation subunit is used to perform simulation and economic and safety evaluation on the generated scheduling strategy to select the optimal strategy.

[0022] The intelligent prediction module includes a historical data analysis subunit, a model training subunit and a prediction result correction subunit, and the historical data analysis subunit, the model training subunit and the prediction result correction subunit are all connected to the user interaction module;

[0023] The historical data analysis subunit is used to mine and analyze historical energy data and extract potential patterns;

[0024] The model training subunit is used to train the prediction model using a deep learning framework;

[0025] The prediction result correction subunit is used to correct the prediction result by combining real-time data with expert experience.

[0026] Wherein, the user interaction module includes an information display subunit, a demand response subunit and a service customization subunit;

[0027] The information display subunit is used to display the user's energy consumption, profit analysis and the overall operation status of the virtual power plant in the form of charts, reports, etc.

[0028] The demand response subunit is used to receive the user's demand response application and adjust the optimization scheduling strategy according to the application;

[0029] The service customization subunit is used to allow users to customize energy service packages according to their own needs.

[0030] The self-learning adaptive module includes a performance monitoring subunit, a model updating subunit and a parameter optimization subunit;

[0031] The performance monitoring subunit is used to monitor the operating indicators of the virtual power plant in real time;

[0032] The model updating subunit is used to update the optimization scheduling model and prediction model using online learning or transfer learning algorithms according to the performance monitoring results;

[0033] The parameter optimization subunit is used to fine-tune the model parameters using an evolutionary algorithm or other optimization algorithms.

[0034] The present invention also provides a virtual power plant control method, which is applied to the virtual power plant control system described above and includes the following steps:

[0035] The data fusion module collects energy data from various sensors, electricity meters, and third-party platforms, performs denoising, normalization, and fills missing values in the preprocessing subunit, and then extracts key features and reduces dimensionality in the feature engineering subunit to construct a unified energy data model.

[0036] Based on the intelligent prediction module, the historical data analysis subunit first mines the patterns of historical energy data, then the model training subunit trains the prediction model using a deep learning framework, and finally the prediction result correction subunit corrects the prediction result by combining real-time data and expert experience;

[0037] Based on the environmental perception subunit of the dynamic optimization scheduling module, the strategy generation subunit monitors the power grid status, market prices and policy information in real time. Based on this, the strategy generation subunit uses a deep reinforcement learning algorithm to generate scheduling strategies for different times and scenarios. The strategy evaluation subunit simulates and evaluates the strategies to select the optimal strategy for collaborative control.

[0038] The collaborative control module then decomposes the optimal scheduling strategy into specific control instructions, which are sent to each distributed energy device via a reliable communication protocol. After the device executes the instructions, it will feedback the operating status, and the collaborative control module will monitor the execution status.

[0039] The performance monitoring subunit based on the self-learning adaptive module monitors the virtual power plant operating indicators in real time, the model updating subunit updates the model according to the monitoring results, and the parameter optimization subunit fine-tunes the parameters. At the same time, the user interaction module displays information to the user and receives user demand response applications and service customization requirements;

[0040] Finally, based on the security protection module and the storage unit, the data is protected and stored on the network by deploying firewalls, IDS, and IPS.

[0041] A virtual power plant control system and method of the present invention includes a data fusion module, a dynamic optimization scheduling module, an intelligent prediction module, a user interaction module, a self-learning adaptive module, a collaborative control module, a safety protection module and a storage unit. In terms of data processing, the data fusion module of this design can effectively process massive multi-source heterogeneous data and accurately extract key information through data acquisition, preprocessing and feature engineering; the dynamic optimization scheduling module is based on a deep reinforcement learning algorithm, fully considering multiple factors to generate a global optimal strategy and overcome the drawbacks of local optimization; the safety protection module constructs a multi-layer protection system to ensure system security; the user interaction module covers functions such as information display, demand response and service customization to meet the personalized needs of users. The overall module can cope with the complex situation of energy market and power grid operation, improve the operating efficiency and safety of virtual power plants, and facilitate promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 It is a schematic diagram of the principle of the virtual power plant control system of the present invention.

[0044] Figure 2 It is a schematic diagram of the principle of the data fusion module in the virtual power plant control system of the present invention.

[0045] Figure 3 It is a schematic diagram of the principle of the dynamic optimization scheduling module in the virtual power plant control system of the present invention.

[0046] Figure 4 It is a schematic diagram of the principle of the intelligent prediction module in the virtual power plant control system of the present invention.

[0047] Figure 5 It is a schematic diagram of the principle of the user interaction module in the virtual power plant control system of the present invention.

[0048] Figure 6 It is a schematic diagram of the principle of the self-learning adaptive module in the virtual power plant control system of the present invention.

[0049] Figure 7 It is a schematic diagram of the principle of the collaborative control module in the virtual power plant control system of the present invention.

[0050] Figure 8 It is a schematic diagram of the principle of the safety protection module in the virtual power plant control system of the present invention.

[0051] Figure 9 Schematic diagram of the principles of the modular architecture design module in the virtual power plant control system of the present invention.

[0052] Figure 10 It is a flow chart of the steps of a virtual power plant control method of the present invention.

[0053] 1-Data fusion module, 2-Dynamic optimization scheduling module, 3-Intelligent prediction module, 4-User interaction module, 5-Self-learning and adaptive module, 6-Cooperative control module, 7-Safety protection module, 8-Storage unit, 9-Data acquisition subunit, 10-Data preprocessing subunit, 11-Feature engineering subunit, 12-Environmental perception subunit, 13-Strategy generation subunit, 14-Strategy evaluation subunit, 15-Historical data analysis subunit, 16-Model training subunit, 17-Prediction result correction subunit, 18- Information display subunit, 19-demand response subunit, 20-service customization subunit, 21-performance monitoring subunit, 22-model update subunit, 23-parameter optimization subunit, 24-instruction distribution subunit, 25-device communication subunit, 26-execution feedback subunit, 27-network security protection subunit, 28-data encryption subunit, 29-device fault diagnosis subunit, 30-modular architecture design module, 31-service splitting subunit, 32-containerized deployment subunit, 33-service governance subunit. DETAILED DESCRIPTION

[0054] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0055] See also Figures 1 to 9 The present invention provides a virtual power plant control system, comprising a data fusion module 1, a dynamic optimization scheduling module 2, an intelligent prediction module 3, and a user interaction module 4. The data fusion module 1 is connected to the dynamic optimization scheduling module 2, the dynamic optimization scheduling module 2 is connected to the intelligent prediction module 3, the intelligent prediction module 3 is connected to the user interaction module 4, the dynamic optimization scheduling module 2 and the intelligent prediction module 3 are connected to a self-learning adaptive module 5, the dynamic optimization scheduling module 2 is connected to a collaborative control module 6, and the data fusion module 1, the dynamic optimization scheduling module 2, the intelligent prediction module 3, and the user interaction module 4 are connected to a security protection module 7 and a storage unit 8;

[0056] The data fusion module 1 is used to collect and integrate multi-source heterogeneous data from different types of distributed energy terminals, power grid monitoring equipment, and external energy markets, and build a unified energy data model through data cleaning, feature extraction, and fusion processing;

[0057] The dynamic optimization scheduling module 2 is used to generate a global optimal energy scheduling strategy based on the energy data model, combined with real-time electricity price signals, grid load demand, and the operating status and constraints of each distributed energy device using a deep reinforcement learning algorithm;

[0058] The intelligent prediction module 3 is used to use deep learning neural networks to make high-precision predictions of distributed energy power output, load change trends, and market price fluctuations;

[0059] The user interaction module 4 is used to provide a visual user interface;

[0060] The self-learning and adaptive module 5 is used to automatically adjust the parameters of the optimization scheduling model and the prediction model according to the data feedback and performance evaluation results during the actual operation of the virtual power plant;

[0061] The collaborative control module 6 is used to decompose the strategy of the dynamic optimization scheduling module 2 into specific control instructions and send them to each distributed energy device through a reliable communication protocol;

[0062] The safety protection module 7 is used to build a multi-layer safety protection system;

[0063] The storage unit 8 is used to store implementation data.

[0064] In this embodiment, the data fusion module 1 is used to collect and integrate multi-source heterogeneous data from different types of distributed energy terminals, power grid monitoring equipment, and external energy markets, and construct a unified energy data model through data cleaning, feature extraction, and fusion processing. The dynamic optimization scheduling module 2 is used to generate a globally optimal energy scheduling strategy based on the energy data model, combined with real-time electricity price signals, power grid load demand, and the operating status and constraints of each distributed energy device using a deep reinforcement learning algorithm. The intelligent prediction module 3 is used to use a deep learning neural network to perform high-precision predictions of the power output, load change trends, and market price fluctuations of distributed energy. The user interaction module 4 is used to provide a visual user interface. The self-learning and adaptive module 5 is used to automatically adjust the parameters of the optimization scheduling model and the prediction model based on data feedback and performance evaluation results during the actual operation of the virtual power plant. The collaborative control module 6 is used to decompose the strategy of the dynamic optimization scheduling module 2 into specific control instructions and issue them to each distributed energy device through a reliable communication protocol. The security protection module 7 is used to build a multi-layer security protection system. The storage unit 8 is used to store implementation data.

[0065] Furthermore, the data fusion module 1 includes a data acquisition subunit 9, a data preprocessing subunit 10 and a feature engineering subunit 11, and the data acquisition subunit 9, the data preprocessing subunit 10 and the feature engineering subunit 11 are all connected to the dynamic optimization scheduling module 2;

[0066] The data acquisition subunit 9 is responsible for collecting energy data from various sensors, smart meters and third-party data platforms;

[0067] The data preprocessing subunit 10 is used to perform denoising, normalization, and missing value filling on the collected data;

[0068] The feature engineering subunit 11 is used to extract key features from the data and perform dimensionality reduction.

[0069] In this embodiment, the data acquisition subunit 9 is responsible for collecting energy data from various sensors, smart meters and third-party data platforms; the data preprocessing subunit 10 is used to denoise, normalize and fill missing values in the collected data; the feature engineering subunit 11 is used to extract key features in the data and perform dimensionality reduction.

[0070] Furthermore, the dynamic optimization scheduling module 2 includes an environment perception subunit 12, a strategy generation subunit 13 and a strategy evaluation subunit 14, and the environment perception subunit 12, the strategy generation subunit 13 and the strategy evaluation subunit 14 are all connected to the intelligent prediction module 3;

[0071] The environmental sensing subunit 12 is used to monitor the power grid operation status, energy market prices and policy information in real time;

[0072] The strategy generation subunit 13 is used to generate scheduling strategies for different time scales and scenarios based on the model constructed by the deep reinforcement learning algorithm;

[0073] The strategy evaluation subunit 14 is used to perform simulation and economic and safety evaluation on the generated scheduling strategy to select the optimal strategy.

[0074] In this embodiment, the environmental perception subunit 12 is used to monitor the operating status of the power grid, energy market prices and policy information in real time; the strategy generation subunit 13 is used to generate scheduling strategies for different time scales and scenarios based on the model constructed by the deep reinforcement learning algorithm; the strategy evaluation subunit 14 is used to simulate and evaluate the economy and safety of the generated scheduling strategy to screen out the optimal strategy.

[0075] Furthermore, the intelligent prediction module 3 includes a historical data analysis subunit 15, a model training subunit 16 and a prediction result correction subunit 17, and the historical data analysis subunit 15, the model training subunit 16 and the prediction result correction subunit 17 are all connected to the user interaction module 4;

[0076] The historical data analysis subunit 15 is used to mine and analyze historical energy data and extract potential patterns;

[0077] The model training subunit 16 is used to train the prediction model using a deep learning framework;

[0078] The prediction result correction subunit 17 is used to correct the prediction result by combining real-time data with expert experience.

[0079] In this embodiment, the historical data analysis subunit 15 is used to mine and analyze historical energy data to extract potential patterns; the model training subunit 16 is used to train the prediction model using a deep learning framework; and the prediction result correction subunit 17 is used to correct the prediction results by combining real-time data and expert experience.

[0080] Furthermore, the user interaction module 4 includes an information display subunit 18, a demand response subunit 19 and a service customization subunit 20;

[0081] The information display subunit 18 is used to display the user's energy consumption, profit analysis and the overall operation status of the virtual power plant in the form of charts, reports, etc.

[0082] The demand response subunit 19 is used to receive the user's demand response application and adjust the optimization scheduling strategy according to the application;

[0083] The service customization subunit 20 is used to allow users to customize energy service packages according to their own needs.

[0084] In this embodiment, the information display subunit 18 is used to display the user's energy consumption, profit analysis and the overall operating status of the virtual power plant in the form of charts, reports, etc.; the demand response subunit 19 is used to receive the user's demand response application and adjust the optimization scheduling strategy according to the application; the service customization subunit 20 is used to allow the user to customize the energy service package according to his or her own needs.

[0085] Furthermore, the self-learning adaptive module 5 includes a performance monitoring subunit 21, a model updating subunit 22 and a parameter optimization subunit 23;

[0086] The performance monitoring subunit 21 is used to monitor the operating indicators of the virtual power plant in real time;

[0087] The model updating subunit 22 is used to update the optimization scheduling model and prediction model using online learning or transfer learning algorithms according to the performance monitoring results;

[0088] The parameter optimization subunit 23 is used to fine-tune the model parameters using an evolutionary algorithm or other optimization algorithms.

[0089] In this embodiment, the performance monitoring subunit 21 is used to monitor the operating indicators of the virtual power plant in real time; the model updating subunit 22 is used to update the optimization scheduling model and prediction model based on the performance monitoring results using online learning or transfer learning algorithms; the parameter optimization subunit 23 is used to fine-tune the model parameters using evolutionary algorithms or other optimization algorithms.

[0090] Furthermore, the collaborative control module 6 includes an instruction distribution subunit 24, a device communication subunit 25 and an execution feedback subunit 26;

[0091] The instruction distribution subunit 24 is used to convert the optimized scheduling strategy into specific control instructions and encapsulate them according to the device communication protocol;

[0092] The device communication subunit 25 is used to reliably transmit control instructions to each distributed energy device through power line communication, wireless communication or optical fiber communication;

[0093] The execution feedback subunit 26 is used to receive execution feedback information from each device to ensure effective execution of the control instructions.

[0094] In this embodiment, the instruction distribution subunit 24 is used to convert the optimized scheduling strategy into specific control instructions and encapsulate them according to the device communication protocol; the device communication subunit 25 is used to reliably transmit the control instructions to each distributed energy device through power line communication, wireless communication or optical fiber communication; the execution feedback subunit 26 is used to receive execution feedback information from each device to ensure the effective execution of the control instructions.

[0095] Furthermore, the security protection module 7 includes a network security protection subunit 27, a data encryption subunit 28 and an equipment fault diagnosis subunit 29;

[0096] The network security protection subunit 27 is used to deploy firewalls, intrusion detection systems and intrusion prevention systems to prevent network attacks;

[0097] The data encryption subunit 28 is used to encrypt data transmission and storage using a combination of symmetric encryption and asymmetric encryption;

[0098] The equipment fault diagnosis subunit 29 is used to diagnose equipment faults in real time based on equipment status monitoring data using a fault diagnosis algorithm, and to perform isolation and recovery processing.

[0099] In this embodiment, the network security protection subunit 27 is used to deploy firewalls, intrusion detection systems and intrusion prevention systems to prevent network attacks; the data encryption subunit 28 is used to encrypt data transmission and storage by combining symmetric encryption and asymmetric encryption; the equipment fault diagnosis subunit 29 is used to diagnose equipment faults in real time based on equipment status monitoring data using a fault diagnosis algorithm, and to perform isolation and recovery processing.

[0100] Furthermore, the virtual power plant control system also includes a modular architecture design module 30, which is connected to the user interaction module 4, and the modular architecture design module 30 includes a service splitting subunit 31, a containerized deployment subunit 32 and a service governance subunit 33.

[0101] The service splitting subunit 31 is used to split the system functions into multiple independent microservices according to the business logic

[0102] The containerized deployment subunit 32 is used to encapsulate and deploy microservices using container technology such as Docker;

[0103] The service governance subunit 33 is used to implement registration and discovery, load balancing, configuration management, and monitoring and operation of microservices.

[0104] In this embodiment, the service splitting subunit 31 is used to split the system functions into multiple independent microservices according to business logic; the containerized deployment subunit 32 is used to encapsulate and deploy microservices using container technology such as Docker; the service governance subunit 33 is used to implement registration and discovery, load balancing, configuration management and monitoring and operation of microservices.

[0105] Based on the present invention, please refer to Figure 10 The present invention also provides a virtual power plant control method, which is applied to the virtual power plant control system described above and includes the following steps:

[0106] S1: Energy data is collected from various sensors, electricity meters, and third-party platforms through the data fusion module 1. The preprocessing subunit performs denoising, normalization, and missing value filling. The feature engineering subunit 11 then extracts key features and reduces the dimensionality to construct a unified energy data model.

[0107] S2: Based on the intelligent prediction module 3, the historical data analysis subunit 15 first mines the patterns of historical energy data, then the model training subunit 16 trains the prediction model using a deep learning framework, and finally the prediction result correction subunit 17 corrects the prediction result by combining real-time data and expert experience;

[0108] S3: Based on the real-time monitoring of the grid status, market prices and policy information by the environment perception subunit 12 of the dynamic optimization scheduling module 2, the strategy generation subunit 13 generates scheduling strategies for different times and scenarios using a deep reinforcement learning algorithm. The strategy evaluation subunit 14 simulates and evaluates the strategies to select the optimal strategy for collaborative control.

[0109] S4: The optimal scheduling strategy is then decomposed into specific control instructions by the collaborative control module 6, which are sent to each distributed energy device via a reliable communication protocol. After the device executes the instruction, it feedbacks the operating status, and the collaborative control module 6 monitors the execution status.

[0110] S5: The performance monitoring subunit 21 of the self-learning adaptive module 5 monitors the virtual power plant operating indicators in real time, the model updating subunit 22 updates the model according to the monitoring results, and the parameter optimization subunit 23 fine-tunes the parameters. At the same time, the user interaction module 4 displays information to the user and receives user demand response applications and service customization requirements;

[0111] S6: Finally, based on the security protection module 7 and the storage unit 8, the data is protected and stored through the deployment of firewalls, IDS, and IPS.

[0112] The above disclosure is only a preferred embodiment of the present invention, and certainly cannot be used to limit the scope of the rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made in accordance with the claims of the present invention are still within the scope of the invention.

Claims

1. A virtual power plant control system, characterized in that: It includes a data fusion module, a dynamic optimization scheduling module, an intelligent prediction module and a user interaction module. The data fusion module is connected to the dynamic optimization scheduling module, the dynamic optimization scheduling module is connected to the intelligent prediction module, and the intelligent prediction module is connected to the user interaction module. The dynamic optimization scheduling module and the intelligent prediction module are connected to a self-learning adaptive module, the dynamic optimization scheduling module is connected to a collaborative control module, and the data fusion module, the dynamic optimization scheduling module, the intelligent prediction module, and the user interaction module are connected to a security protection module and a storage unit; The data fusion module is used to collect and integrate multi-source heterogeneous data from different types of distributed energy terminals, power grid monitoring equipment, and external energy markets, and build a unified energy data model through data cleaning, feature extraction, and fusion processing; The dynamic optimization scheduling module is used to generate a global optimal energy scheduling strategy based on the energy data model, combined with real-time electricity price signals, grid load demand, and the operating status and constraints of each distributed energy device using a deep reinforcement learning algorithm; The intelligent prediction module is used to use deep learning neural networks to make high-precision predictions of distributed energy power output, load change trends, and market price fluctuations; The user interaction module is used to provide a visual user interface; The self-learning and adaptive module is used to automatically adjust the parameters of the optimization scheduling model and the prediction model according to the data feedback and performance evaluation results during the actual operation of the virtual power plant; The collaborative control module is used to decompose the strategy of the dynamic optimization scheduling module into specific control instructions and send them to each distributed energy device through a reliable communication protocol; The security protection module is used to build a multi-layer security protection system; The storage unit is used to store implementation data.

2. The virtual power plant control system according to claim 1, characterized in that: The data fusion module includes a data acquisition subunit, a data preprocessing subunit and a feature engineering subunit, and the data acquisition subunit, the data preprocessing subunit and the feature engineering subunit are all connected to the dynamic optimization scheduling module; The data acquisition subunit is responsible for collecting energy data from various sensors, smart meters and third-party data platforms; The data preprocessing subunit is used to perform denoising, normalization, and missing value filling on the collected data; The feature engineering subunit is used to extract key features from the data and perform dimensionality reduction.

3. The virtual power plant control system according to claim 2, characterized in that: The dynamic optimization scheduling module includes an environment perception subunit, a strategy generation subunit and a strategy evaluation subunit, and the environment perception subunit, the strategy generation subunit and the strategy evaluation subunit are all connected to the intelligent prediction module; The environmental sensing subunit is used to monitor the power grid operation status, energy market prices and policy information in real time; The strategy generation subunit is used to generate scheduling strategies for different time scales and scenarios based on the model constructed by the deep reinforcement learning algorithm; The strategy evaluation subunit is used to perform simulation and economic and safety evaluation on the generated scheduling strategy to select the optimal strategy.

4. The virtual power plant control system according to claim 3, characterized in that: The intelligent prediction module includes a historical data analysis subunit, a model training subunit and a prediction result correction subunit, and the historical data analysis subunit, the model training subunit and the prediction result correction subunit are all connected to the user interaction module; The historical data analysis subunit is used to mine and analyze historical energy data and extract potential patterns; The model training subunit is used to train the prediction model using a deep learning framework; The prediction result correction subunit is used to correct the prediction result by combining real-time data with expert experience.

5. The virtual power plant control system according to claim 4, characterized in that: The user interaction module includes an information display subunit, a demand response subunit, and a service customization subunit; The information display subunit is used to display the user's energy consumption, profit analysis and the overall operation status of the virtual power plant in the form of charts, reports, etc. The demand response subunit is used to receive the user's demand response application and adjust the optimization scheduling strategy according to the application; The service customization subunit is used to allow users to customize energy service packages according to their own needs.

6. The virtual power plant control system according to claim 5, characterized in that: The self-learning adaptive module includes a performance monitoring subunit, a model updating subunit and a parameter optimization subunit; The performance monitoring subunit is used to monitor the operating indicators of the virtual power plant in real time; The model updating subunit is used to update the optimization scheduling model and prediction model using online learning or transfer learning algorithms according to the performance monitoring results; The parameter optimization subunit is used to fine-tune the model parameters using an evolutionary algorithm or other optimization algorithms.

7. A virtual power plant control method, applied to the virtual power plant control system according to claim 6, characterized in that: The steps include: The data fusion module collects energy data from various sensors, electricity meters, and third-party platforms, performs denoising, normalization, and fills missing values in the preprocessing subunit, and then extracts key features and reduces dimensionality in the feature engineering subunit to construct a unified energy data model. Based on the intelligent prediction module, the historical data analysis subunit first mines the patterns of historical energy data, then the model training subunit trains the prediction model using a deep learning framework, and finally the prediction result correction subunit corrects the prediction result by combining real-time data and expert experience; Based on the environmental perception subunit of the dynamic optimization scheduling module, the strategy generation subunit monitors the power grid status, market prices and policy information in real time. Based on this, the strategy generation subunit uses a deep reinforcement learning algorithm to generate scheduling strategies for different times and scenarios. The strategy evaluation subunit simulates and evaluates the strategies to select the optimal strategy for collaborative control. The collaborative control module then decomposes the optimal scheduling strategy into specific control instructions, which are sent to each distributed energy device via a reliable communication protocol. After the device executes the instructions, it will feedback the operating status, and the collaborative control module will monitor the execution status. The performance monitoring subunit based on the self-learning adaptive module monitors the virtual power plant operating indicators in real time, the model updating subunit updates the model according to the monitoring results, and the parameter optimization subunit fine-tunes the parameters. At the same time, the user interaction module displays information to the user and receives user demand response applications and service customization requirements; Finally, based on the security protection module and the storage unit, the data is protected and stored on the network by deploying firewalls, IDS, and IPS.

Images