Intelligent virtual power plant cross-domain collaborative optimization system and method

Abstract

The invention discloses an intelligent virtual power plant cross-domain collaborative optimization system. The system comprises a data acquisition unit which comprises an energy controller, an intelligent electric meter, a sensor and an edge gateway; the edge calculation unit comprises a federated learning optimization module, and the federated learning optimization module is used for carrying out federated learning optimization; based on a lightweight federated learning node local training model, adopting a differential privacy technology to desensitize data; the cloud collaboration unit comprises a multi-modal large model, and the multi-modal large model is used for aggregating edge model parameters and generating a global optimization strategy; and the execution unit comprises an intelligent contract. According to the method, global optimization of industrial, commercial and resident resources is realized through cross-domain collaborative scheduling, and compared with existing single-group scheduling, the efficiency of the system is improved; the communication overhead is reduced by balancing privacy and efficiency, locally processing sensitive data by the edge computing layer and only transmitting model parameters by the cloud, and the system response time is greatly shortened by executing real-time dynamic response and combining reinforcement learning dynamic adjustment strategies.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual power plants, and specifically to an intelligent virtual power plant cross-domain collaborative optimization system and method. Background Art

[0002] The current virtual power plant technology faces the following challenges:

[0003] 1. Data island and privacy issues: The electricity consumption data of industrial, commercial, and residential users is scattered and sensitive, and it is difficult for traditional centralized models to achieve cross-domain sharing and collaborative optimization.

[0004] 2. Insufficient adaptability to complex scenarios: The output of renewable energy is highly volatile, and traditional scheduling models rely on a single data source, making it difficult to accurately predict.

[0005] 3. Low real-time response efficiency: Existing methods such as genetic algorithms or particle swarm optimization have high computational complexity and are difficult to meet the second-level scheduling requirements.

[0006] Therefore, we propose to design an intelligent virtual power plant cross-domain collaborative optimization system and method. Summary of the Invention

[0007] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this part, the abstract, and the title. However, such simplifications or omissions shall not be used to limit the scope of the present invention.

[0008] To solve the above technical problems, according to one aspect of the present invention, the following technical solutions are provided:

[0009] An intelligent virtual power plant cross-domain collaborative optimization system, comprising:

[0010] A data acquisition unit: including an energy controller, an intelligent electricity meter, sensors, and an edge gateway;

[0011] An edge computing unit: including a federated learning optimization module, and the federated learning optimization module is used for desensitizing data;

[0012] A cloud collaboration unit: including a multi-modal large model, and the multi-modal large model is used for aggregating edge model parameters and generating a global optimization strategy;

[0013] An execution unit: including an intelligent contract, and the intelligent contract is used for automatically executing scheduling instructions.

[0014] As a preferred solution of the intelligent virtual power plant cross - domain collaborative optimization system described in the present invention, the multimodal large model includes a Transformer architecture. By using the Transformer architecture, the edge computing layer locally processes sensitive data, and only model parameters are transmitted to the cloud, greatly reducing communication overhead.

[0015] As a preferred solution of the intelligent virtual power plant cross - domain collaborative optimization system described in the present invention, the dispatching instructions include energy storage charge - discharge instructions and demand - response load adjustment instructions. Through smart contracts deployed on the blockchain network, when preset dispatching conditions are met, the smart contracts will automatically execute corresponding operations to ensure the transparency and immutability of instruction execution.

[0016] The intelligent virtual power plant cross - domain collaborative optimization method includes the following steps:

[0017] Step 1: Deploy energy controllers, smart meters, sensors, and edge gateways to collect distributed energy, load, and environmental data.

[0018] Step 2: Locally train the model based on lightweight federated learning nodes, desensitize data using differential privacy technology. In the federated learning optimization, the edge nodes use LSTM to predict local load, and the cloud fuses model parameters through an adaptive aggregation algorithm to avoid data leakage. Introduce a GAN (Generative Adversarial Network) to simulate extreme scenarios, such as a sudden drop in photovoltaic power output caused by a typhoon, etc., to enhance the robustness of the model.

[0019] Step 3: Integrate the multimodal large model, aggregate edge model parameters, generate a global optimization strategy, construct a cross - modal feature extractor based on Transformer, and integrate power data, meteorological information, and market trading signals to improve prediction accuracy.

[0020] Step 4: Automatically execute dispatching instructions through smart contracts.

[0021] As a preferred solution of the intelligent virtual power plant cross - domain collaborative optimization method described in the present invention, the distributed energy includes photovoltaic energy, wind power generation energy, natural gas power generation energy, and fuel cell energy storage.

[0022] As a preferred solution of the intelligent virtual power plant cross - domain collaborative optimization method described in the present invention, the environmental data includes air temperature, humidity, and weather conditions.

[0023] Compared with the prior art, the beneficial effects of the present invention are:

[0024] Through cross - domain collaborative scheduling, the present invention realizes the global optimization of industrial, commercial and residential resources, improving the system efficiency compared with the existing single - group scheduling; by balancing privacy and efficiency, the edge computing layer locally processes sensitive data, and only model parameters are transmitted to the cloud, reducing the communication overhead; by implementing real - time dynamic response and dynamically adjusting strategies in combination with reinforcement learning, the system response time is greatly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the present invention will be described in detail below with reference to the drawings and specific embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts. Among them:

[0026] Figure 1 It is a system diagram of the cross - domain collaborative optimization system and method of the intelligent virtual power plant of the present invention; SPECIFIC EMBODIMENTS

[0027] In order to make the above - mentioned objects, features and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention will be described in detail below with reference to the drawings.

[0028] Please refer to Figure 1 , the present invention provides an intelligent virtual power plant cross - domain collaborative optimization system, including:

[0029] Data acquisition unit: including an energy controller, an intelligent electricity meter, sensors and an edge gateway;

[0030] Edge computing unit: including a federated learning optimization module, and the federated learning optimization module is used to desensitize data;

[0031] Cloud collaborative unit: including a multi - modal large model, and the multi - modal large model is used to aggregate edge model parameters and generate a global optimization strategy;

[0032] Among them, the multi - modal large model includes a Transformer architecture. By adopting the Transformer architecture, the edge computing layer locally processes sensitive data, and only model parameters are transmitted to the cloud, greatly reducing the communication overhead.

[0033] Execution unit: including an intelligent contract, and the intelligent contract is used to automatically execute scheduling instructions. The scheduling instructions include energy storage charge - discharge instructions and demand response load adjustment instructions. Through the intelligent contract deployed on the blockchain network, when the preset scheduling conditions are met, the intelligent contract will automatically execute the corresponding operations to ensure the transparency and immutability of the instruction execution.

[0034] The cross - domain collaborative optimization method of the intelligent virtual power plant includes the following steps:

[0035] Step 1: Deploy energy controllers, smart meters, sensors, and edge gateways to collect distributed energy, load, and environmental data.

[0036] Among them, distributed energy includes photovoltaic energy, wind power generation energy, natural gas power generation energy, and fuel cell energy storage. Environmental data includes air temperature, humidity, and weather conditions.

[0037] Step 2: Based on the lightweight federated learning node local training model, such as industrial areas and commercial buildings, use differential privacy technology to desensitize data. In the federated learning optimization, the edge node uses LSTM to predict the local load, and the cloud fuses the model parameters through an adaptive aggregation algorithm to avoid data leakage. Introduce the GAN (Generative Adversarial Network) to simulate extreme scenarios, such as a sharp drop in photovoltaic output caused by a typhoon, etc., to enhance the robustness of the model.

[0038] Step 3: Integrate a multi-modal large model, aggregate the edge model parameters, generate a global optimization strategy, construct a cross-modal feature extractor based on Transformer, and integrate power data, meteorological information, and market trading signals to improve the prediction accuracy.

[0039] Step 4: Automatically execute scheduling instructions through smart contracts.

[0040] Although the present invention has been described above with reference to the embodiments, various improvements can be made without departing from the scope of the present invention. The features of the embodiments disclosed in the present invention can be combined with each other in any way. The exhaustive description of these combinations is not given in this specification only for the sake of saving space and resources. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. An intelligent virtual power plant cross - domain collaborative optimization system, characterized in that Including: Data acquisition unit: including an energy controller, a smart meter, sensors and an edge gateway; Edge computing unit: including a federated learning optimization module for desensitizing data; Cloud collaboration unit: including a multi-modal large model for aggregating edge model parameters and generating a global optimization strategy; Execution unit: including a smart contract for automatically executing scheduling instructions.

2. The intelligent virtual power plant cross-domain collaborative optimization system according to claim 1, characterized in that, The multi-modal large model includes a Transformer architecture.

3. The intelligent virtual power plant cross-domain collaborative optimization system according to claim 1, wherein The scheduling instructions include energy storage charge and discharge instructions and demand response load adjustment instructions.

4. The intelligent virtual power plant cross-domain collaborative optimization method according to any one of claims 1-3, characterized in that Including the following steps: Step 1: Deploy an energy controller, a smart meter, sensors and an edge gateway to collect distributed energy, load and environmental data; Step 2: Based on a lightweight federated learning node local training model, use differential privacy technology to desensitize data. In federated learning optimization, the edge node uses LSTM to predict the local load, and the cloud fuses the model parameters through an adaptive aggregation algorithm to avoid data leakage. Introduce a GAN generative adversarial network to simulate extreme scenarios and enhance the robustness of the model; Step 3: Integrate a multi-modal large model, aggregate edge model parameters, generate a global optimization strategy, build a cross-modal feature extractor based on Transformer, integrate power data, meteorological information and market trading signals to improve prediction accuracy; Step 4: Automatically execute scheduling instructions through a smart contract.

5. The intelligent virtual power plant cross-domain collaborative optimization method according to claim 4, wherein The distributed energy includes photovoltaic energy, wind power energy, natural gas power generation energy and fuel cell energy storage.

6. The intelligent virtual power plant cross-domain collaborative optimization method according to claim 4, characterized in that The environmental data includes air temperature, humidity and weather conditions.

Images