Micro-grid system energy management optimization method and system based on big data analysis

Abstract

The invention belongs to the technical field of energy management optimization, and discloses a micro-grid system energy management optimization method and system based on big data analysis. The method comprises the following steps: constructing a multi-source time sequence data processing model, an energy supply and demand prediction model and an energy management optimization model of a micro-grid system; updating the model according to an alliance learning mechanism; performing data processing on the real-time multi-source time sequence data; acquiring a real-time topology network and a dynamic data model; energy supply and demand prediction is carried out, and an energy management optimization scheme is generated; and visualizing the real-time energy supply and demand prediction result and the real-time energy management optimization scheme. The problems that in the prior art, the data processing capacity is limited, model updating is not timely, real-time performance is lacked, prediction precision is insufficient, and an optimization scheme is not flexible enough are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy management optimization, and specifically relates to a microgrid system energy management optimization method and system based on big data analysis. Background Art

[0002] With the widespread use of renewable energy, microgrid systems, as an important component of smart grids, have become increasingly important for improving energy efficiency and system stability. However, existing technologies still have many drawbacks, including:

[0003] 1) Limited data processing capabilities: Existing technologies often exhibit insufficient data processing capabilities when processing large-scale, multi-source, and heterogeneous time series data, limiting the system's adaptability and response speed to complex energy scenarios.

[0004] 2) Model updates are not timely: Model updates in existing technologies usually require centralized processing, which is not only time-consuming but may not reflect the latest status of the microgrid system in a timely manner, resulting in a decrease in the accuracy of prediction and optimization results;

[0005] 3) Lack of real-time performance: Due to delays in data processing and model updates, existing technologies may not be able to provide real-time energy management optimization solutions, which is a significant drawback in microgrid environments that require fast response.

[0006] 4) Insufficient forecasting accuracy: Existing energy supply and demand forecasting models may not be able to fully utilize the latest data and technologies to make accurate forecasts, especially when faced with complex and changing energy supply and demand scenarios.

[0007] 5) Optimization schemes are not flexible enough: Traditional energy management optimization models are often based on preset optimization objectives and fitness functions. These presets may not be able to adapt to the real-time changes in energy supply and demand, resulting in inflexible and ineffective optimization schemes. Summary of the Invention

[0008] In order to solve the problems of limited data processing capability, untimely model updates, lack of real-time performance, insufficient prediction accuracy and inflexible optimization schemes in the existing technology, the present invention aims to provide a microgrid system energy management optimization method and system based on big data analysis.

[0009] The technical solution adopted in the present invention is:

[0010] A microgrid system energy management optimization method based on big data analysis includes the following steps:

[0011] The cloud data center uses artificial intelligence algorithms to analyze the collected multi-source time series big data and build a multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model for the microgrid system.

[0012] The cloud data center updates the model based on the federated learning mechanism to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model. It also synchronizes the updated multi-source time series data processing model to the edge computing gateways of all monitoring points.

[0013] The edge computing gateway collects real-time multi-source time series data from monitoring points corresponding to the microgrid system, uses the updated multi-source time series data processing model to process the real-time multi-source time series data, obtains real-time standard time series data, and uploads the real-time standard time series data to the cloud data center;

[0014] The cloud data center collects the real-time connection data of the microgrid system, builds the corresponding real-time topology network model based on the real-time connection data of the microgrid system, and adds the real-time standard time series data of all monitoring points to the real-time topology network model to obtain the real-time topology network and dynamic data model;

[0015] The cloud data center uses the updated energy supply and demand forecasting model to perform energy supply and demand forecasting based on the real-time topology network and dynamic data model. Based on the obtained real-time energy supply and demand forecasting results, the updated energy management optimization model is used to generate an energy management optimization solution to obtain a real-time energy management optimization solution.

[0016] The cloud data center uses real-time topology networks and dynamic data models to visualize real-time energy supply and demand forecast results and real-time energy management optimization solutions, and synchronizes the real-time energy management optimization solutions to all microgrid devices in the microgrid system through all edge computing gateways.

[0017] Furthermore, the cloud data center uses artificial intelligence algorithms to analyze the collected multi-source time series big data and build a multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model for the microgrid system, including the following steps:

[0018] The cloud data center collects sample connection data of several microgrid systems and uses topology network simulation technology to build a corresponding sample topology network model based on the sample connection data.

[0019] Collect sample multi-source time series big data including different data sources in the microgrid system, and preprocess them to obtain a number of preprocessed sample multi-source time series data;

[0020] Based on several pre-processed sample multi-source time series data, a multi-source time series data processing model is constructed using deep learning and reinforcement learning algorithms, and several sample standard time series data are obtained;

[0021] Based on several sample topological network models and several sample standard time series data, a deep learning and feature screening algorithm is used to build an energy supply and demand forecasting model and obtain several sample energy supply and demand forecasting results;

[0022] Based on the energy supply and demand forecast results of several samples, an energy management optimization model is constructed using deep learning and swarm intelligence optimization algorithms.

[0023] Furthermore, the multi-source time series data processing model is constructed based on the CNN-LSTM-MOGRPO-DPA algorithm, and the multi-source time series data processing model includes an unstructured data feature extraction module constructed based on the CNN algorithm, a structured data feature extraction module constructed based on the LSTM algorithm, a data processing strategy generation module constructed based on the MOGRPO algorithm, and a data processing algorithm library constructed based on the DPA algorithm, and the data processing algorithm library includes several DPA algorithm packages.

[0024] Furthermore, the energy supply and demand forecasting model is constructed based on the GCN-RF-MLP algorithm, and the energy supply and demand forecasting model includes a graph structure feature extraction module constructed based on the GCN algorithm, a key feature screening module constructed based on the RF algorithm, and an energy supply and demand forecasting module constructed based on the MLP algorithm, which are connected in sequence.

[0025] Furthermore, the energy management optimization model is constructed based on the ISGA algorithm, and the energy management optimization model includes an influence factor generation module, a fitness function update module, an initial solution generation module, an iterative update module and a solution vector analysis module which are connected in sequence.

[0026] Furthermore, the cloud data center updates the model based on the federated learning mechanism to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model. The updated multi-source time series data processing model is then synchronized to the edge computing gateways of all monitoring points, including the following steps:

[0027] The cloud data center builds a corresponding historical topology network model based on the historical connection data of the microgrid system, extracts the initial model parameters of the historical topology network model, multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model, and sends them to all edge computing gateways;

[0028] The edge computing gateway reconstructs and deploys the initial model parameters of the historical topology network model, multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model based on the initial model parameters.

[0029] The edge computing gateway optimizes and trains the multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model based on the historical topology network model and several historical multi-source time series data collected. It extracts the optimized training model parameters after optimization training and uploads them to the cloud data center.

[0030] The cloud data center integrates the optimized training model parameters uploaded by all edge computing gateways to obtain comprehensive model parameters. Based on the comprehensive model parameters, the cloud data center updates the model to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model.

[0031] The cloud data center extracts the updated model parameters of the updated multi-source time series data processing model and sends the updated model parameters to the edge computing gateways of all monitoring points to complete the synchronization of the updated multi-source time series data processing model.

[0032] Furthermore, the edge computing gateway collects real-time multi-source time series data from monitoring points corresponding to the microgrid system, processes the real-time multi-source time series data using the updated multi-source time series data processing model to obtain real-time standard time series data, and uploads the real-time standard time series data to the cloud data center, including the following steps:

[0033] The edge computing gateway collects real-time multi-source time series data from monitoring points corresponding to the microgrid system and inputs the real-time multi-source time series data into the updated multi-source time series data processing model;

[0034] Using the updated unstructured data feature extraction module of the updated multi-source time series data processing model to extract real-time unstructured data features of the real-time multi-source time series data;

[0035] Using the updated structured data feature extraction module of the updated multi-source time series data processing model to extract real-time structured data features of the real-time multi-source time series data;

[0036] Based on the real-time unstructured data characteristics and real-time structured data characteristics, the data processing strategy generation module of the updated multi-source time series data processing model is used to generate a real-time data processing strategy;

[0037] According to the real-time data processing strategy, several target DPA algorithm packages are called in the data processing algorithm library, and according to several target DPA algorithm packages, real-time multi-source time series data are processed to obtain real-time standard time series data, and the real-time standard time series data is uploaded to the cloud data center.

[0038] Furthermore, the cloud data center uses the updated energy supply and demand forecasting model to perform energy supply and demand forecasting based on the real-time topology network and the dynamic data model. Based on the obtained real-time energy supply and demand forecasting results, the updated energy management optimization model is used to generate an energy management optimization solution. The real-time energy management optimization solution includes the following steps:

[0039] Cloud data center,The cloud data center uses the updated graph structure feature extraction module of the updated energy supply and demand prediction model to extract the real-time graph structure features of the real-time topology network and dynamic data model;

[0040] Extracting several real-time key features of the real-time graph structural features using the updated key feature screening module of the updated energy supply and demand forecasting model;

[0041] An updated energy supply and demand forecasting module of the updated energy supply and demand forecasting model is used to forecast energy supply and demand based on a number of real-time key features to obtain a real-time energy supply and demand forecasting result;

[0042] Using the updated impact factor generation module of the updated energy management optimization model, the real-time energy supply and demand forecast results are analyzed to obtain the real-time impact factors, and the real-time optimization goals are set based on the real-time impact factors;

[0043] Using an updated fitness function updating module of the updated energy management optimization model, the preset fitness function of the updated energy management optimization model is updated according to the real-time optimization target to obtain a real-time fitness function;

[0044] Using an updated initial solution generation module of the updated energy management optimization model, the real-time energy management optimization solution is converted into an individual vector of the updated energy management optimization model, and initial solutions are generated based on the individual vectors to obtain a plurality of initial solutions; the initial solutions correspond to the initial real-time energy management optimization solution;

[0045] Iteratively updating a number of initial solutions according to a real-time fitness function using an updated iterative update module of the updated energy management optimization model, and outputting an optimal solution with the best real-time fitness value;

[0046] The updated solution vector parsing module of the updated energy management optimization model is used to perform solution vector parsing on the individual vectors of the optimal solution to obtain the optimal real-time energy management optimization solution.

[0047] A microgrid system energy management optimization system based on big data analysis is used to implement a microgrid system energy management optimization method. The system includes a cloud data center, several edge computing gateways, several time series data acquisition devices and several microgrid devices. The cloud data center is connected to several edge computing gateways respectively, and each edge computing gateway is communicated with the time series data acquisition device and microgrid device of the corresponding monitoring point, and each time series data acquisition device is communicated with the corresponding microgrid device.

[0048] Furthermore, the cloud data center includes a model building unit, a model updating unit, a data adding unit, an energy supply and demand prediction and energy management optimization unit, and a visualization unit, which are connected in sequence.

[0049] The beneficial effects of the present invention are:

[0050] The present invention provides a microgrid system energy management optimization method and system based on big data analysis. By utilizing big data analysis and artificial intelligence algorithms, it can efficiently process large-scale, multi-source, heterogeneous time series data, thereby improving the adaptability and response speed of the microgrid system to complex energy scenarios; with the help of the alliance learning mechanism, real-time updating of the model is achieved, ensuring that the prediction and optimization results can timely reflect the latest status of the microgrid system, and improving the accuracy of the results; through the real-time data processing and optimization scheme generation of the edge computing gateway, the real-time energy management of the microgrid system is achieved, meeting the needs of rapid response; the constructed energy supply and demand prediction model can discover deep features in time series data, improve the prediction accuracy, and perform better in complex and changeable energy supply and demand scenarios; the constructed energy management optimization model generates more flexible and effective energy management optimization schemes based on real-time energy supply and demand prediction results, which can adapt to real-time changes in energy supply and demand; by introducing edge computing gateways for distributed computing, the scalability and robustness of the system are enhanced, which can improve energy utilization efficiency, reduce operating costs and enhance system stability.

[0051] Other beneficial effects of the present invention will be further described in the specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart of the microgrid system energy management optimization method based on big data analysis in the present invention.

[0053] Figure 2 It is a structural block diagram of the microgrid system energy management optimization system based on big data analysis in the present invention. DETAILED DESCRIPTION

[0054] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0055] Example 1:

[0056] like Figure 1 As shown, this embodiment provides a microgrid system energy management optimization method based on big data analysis, including the following steps:

[0057] S1: The cloud data center uses artificial intelligence algorithms to analyze the collected multi-source time series big data and build a multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model for the microgrid system. The steps include:

[0058] S1-1: The cloud data center collects sample connection data of several microgrid systems and uses topology network simulation technology to build a corresponding sample topology network model based on the sample connection data;

[0059] S1-2: Collect sample multi-source time series big data including different data sources in the microgrid system, and preprocess them to obtain a number of preprocessed sample multi-source time series data;

[0060] Multi-source time series data includes unstructured data and structured data. Unstructured data includes environmental data, weather forecast data, user behavior data, etc., while unstructured data includes distributed energy output data, load demand data, energy storage status data, microgrid interaction data, etc.

[0061] S1-3: Based on several preprocessed sample multi-source time series data, use deep learning and reinforcement learning algorithms to build a multi-source time series data processing model and obtain several sample standard time series data;

[0062] The multi-source time series data processing model is constructed based on the Convolutional Neural Network (CNN)-Long Short-Term Memory (LSTM)-Multi-Objective Group Relative Policy Optimization (MOGRPO)-Data Processing Algorithm (DPA) algorithm. The multi-source time series data processing model includes an unstructured data feature extraction module based on the CNN algorithm, a structured data feature extraction module based on the LSTM algorithm, a data processing strategy generation module based on the MOGRPO algorithm, and a data processing algorithm library based on the DPA algorithm. The data processing algorithm library includes several DPA algorithm packages.

[0063] The unstructured data feature extraction module uses a convolutional neural network (CNN) to extract features from unstructured data. Through operations such as convolution and pooling, it automatically learns the spatial characteristics and local structure of the data, efficiently processes unstructured data, extracts features that are useful for microgrid energy management, and enhances the model's ability to understand and process complex unstructured data.

[0064] The structured data feature extraction module uses a long short-term memory (LSTM) network to extract time series features from structured data (such as distributed energy output, load demand, and energy storage status). This module learns the temporal dependencies and long-term patterns of the data, effectively capturing the time series information and dynamic changes in the structured data. This module addresses the vanishing and exploding gradient problems that are common with traditional methods when processing long-sequence data, improving the model's prediction accuracy and robustness for time series data.

[0065] The data processing strategy generation module includes an objective function set, an intelligent agent, a policy network, and an experience replay pool. The objective function set of the data processing strategy generation module can handle multiple conflicting optimization objective functions, such as data integrity, data complexity, data accuracy, etc., and generate a data processing strategy that balances these objectives. The intelligent agent learns historical data processing strategies through the experience replay pool and continuously optimizes its own strategy generation capabilities. The intelligent agent controls the policy network based on the learned experience to generate more effective data processing strategies. The design of the experience replay pool and the intelligent agent enables the model to continuously learn and optimize, thereby improving the quality of strategy generation. The data processing strategy generation module adopts a group exploration method, which can avoid falling into local optimal solutions to a certain extent. The policy network outputs the distribution probability of actions under a given state. The data processing strategy generation module directly updates the policy network through gradients, eliminating the Critic model in traditional reinforcement learning, making the algorithm structure simpler, and generating a data processing strategy that comprehensively considers multiple objectives, thereby improving the comprehensiveness and rationality of decision-making.

[0066] The data processing algorithm library stores and manages several packaged data processing algorithms (DPA). DPA algorithms include data completion processing algorithms, Gaussian denoising processing algorithms, feature format setting algorithms, normalization processing algorithms, format conversion processing algorithms, etc. This library implements modularization and reuse of data processing algorithms, improves development efficiency, facilitates the selection of appropriate processing algorithms based on different data types and requirements, ensures data quality, and provides a reliable foundation for subsequent analysis and optimization.

[0067] S1-4: Based on several sample topological network models and several sample standard time series data, use deep learning and feature screening algorithms to build an energy supply and demand forecasting model and obtain several sample energy supply and demand forecasting results;

[0068] The energy supply and demand forecasting model is built based on the Graph Convolutional Network (GCN)-Random Forest (RF)-Multi-Layer Perceptron (MLP) algorithm. The model includes a graph structure feature extraction module based on the GCN algorithm, a key feature screening module based on the RF algorithm, and an energy supply and demand forecasting module based on the MLP algorithm.

[0069] The graph structure feature extraction module uses a graph convolutional network (GCN) to extract graph structure features from the microgrid system topology information in the topology network and dynamic data model. By learning the relationship between nodes, it captures the spatial dependency and topological structure in the microgrid system topology network, effectively models the complex relationship between nodes, and enhances the interpretability of the prediction model.

[0070] The key feature screening module uses random forest (RF) to evaluate and screen the importance of features extracted from the GCN module. Through ensemble learning of multiple decision trees, it identifies the most critical features for energy supply and demand forecasting, reduces feature dimensions, reduces model complexity, improves computational efficiency, and ensures that the features entering the final prediction module are representative and important.

[0071] The energy supply and demand forecasting module uses a multi-layer perceptron (MLP) to perform nonlinear mapping and combination on the selected key features. Through multi-layer neural network training, it learns and predicts the complex patterns of energy supply and demand. This module fully utilizes the powerful nonlinear modeling capabilities of MLP to improve the accuracy of energy supply and demand forecasts, flexibly adapts to different energy supply and demand scenarios and changing trends, and provides scientific and accurate forecasting support for energy management optimization.

[0072] S1-5: Based on the energy supply and demand forecast results of several samples, use deep learning and swarm intelligence optimization algorithms to build an energy management optimization model;

[0073] The energy management optimization model is constructed based on the Improved Snow Geese Algorithm (ISGA), and includes an influence factor generation module, a fitness function update module, an initial solution generation module, an iterative update module, and a solution vector analysis module connected in sequence.

[0074] The impact factor generation module is used to analyze the real-time energy supply and demand forecast results, obtain the real-time impact factors, and set real-time optimization goals based on the real-time impact factors;

[0075] A fitness function updating module is used to update the preset fitness function of the updated energy management optimization model according to the real-time optimization target to obtain a real-time fitness function;

[0076] An initial solution generation module is used to convert the real-time energy management optimization solution into individual vectors of the updated energy management optimization model, and generate initial solutions based on the individual vectors to obtain several initial solutions;

[0077] An iterative update module is used to iteratively update several initial solutions according to the real-time fitness function and output the optimal solution with the best real-time fitness value;

[0078] Solution vector analysis module, used to perform solution vector analysis on the individual vectors of the optimal solution to obtain the optimal real-time energy management optimization solution;

[0079] S2: The cloud data center updates the model based on the federated learning mechanism to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model. The updated multi-source time series data processing model is then synchronized to the edge computing gateways at all monitoring points, including the following steps:

[0080] S2-1: The cloud data center builds a corresponding historical topology network model based on the historical connection data of the microgrid system, extracts the initial model parameters of the historical topology network model, multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model, and sends them to all edge computing gateways;

[0081] S2-2: The edge computing gateway reconstructs and deploys the initial model parameters for the historical topology network model, multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model based on the initial model parameters.

[0082] S2-3: The edge computing gateway optimizes and trains the multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model based on the historical topology network model and the collected historical multi-source time series data. It extracts the optimized training model parameters after optimization and uploads them to the cloud data center.

[0083] S2-4: The cloud data center integrates the optimized training model parameters uploaded by all edge computing gateways to obtain comprehensive model parameters. Based on the comprehensive model parameters, the model is updated to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model.

[0084] S2-5: The cloud data center extracts the updated model parameters of the updated multi-source time series data processing model and sends the updated model parameters to the edge computing gateways of all monitoring points to complete the synchronization of the updated multi-source time series data processing model;

[0085] S3: The edge computing gateway collects real-time multi-source time series data from monitoring points corresponding to the microgrid system, uses the updated multi-source time series data processing model to process the real-time multi-source time series data, obtains real-time standard time series data, and uploads the real-time standard time series data to the cloud data center, including the following steps:

[0086] S3-1: Edge computing gateway, collects real-time multi-source time series data from monitoring points corresponding to the microgrid system, and inputs the real-time multi-source time series data into the updated multi-source time series data processing model;

[0087] S3-2: Use the updated unstructured data feature extraction module of the updated multi-source time series data processing model to extract real-time unstructured data features of the real-time multi-source time series data;

[0088] S3-3: Use the updated structured data feature extraction module of the updated multi-source time series data processing model to extract real-time structured data features of the real-time multi-source time series data;

[0089] S3-4: Generate a real-time data processing strategy using the data processing strategy generation module of the updated multi-source time series data processing model based on the real-time unstructured data characteristics and real-time structured data characteristics;

[0090] S3-5: Based on the real-time data processing strategy, call several target DPA algorithm packages in the data processing algorithm library, and process the real-time multi-source time series data according to the target DPA algorithm packages to obtain real-time standard time series data, and upload the real-time standard time series data to the cloud data center;

[0091] S4: The cloud data center collects the real-time connection data of the microgrid system, builds a corresponding real-time topology network model based on the real-time connection data of the microgrid system, and adds the real-time standard time series data of all monitoring points to the real-time topology network model to obtain the real-time topology network and dynamic data model;

[0092] S5: The cloud data center uses the updated energy supply and demand forecasting model to perform energy supply and demand forecasting based on the real-time network topology and dynamic data model. Based on the obtained real-time energy supply and demand forecasting results, the updated energy management optimization model is used to generate an energy management optimization solution. The real-time energy management optimization solution includes the following steps:

[0093] S5-1: Cloud Data Center,The cloud data center uses the updated graph structure feature extraction module of the updated energy supply and demand prediction model to extract the real-time graph structure features of the real-time topology network and dynamic data model;

[0094] S5-2: Using the updated key feature screening module of the updated energy supply and demand forecasting model, extract several real-time key features of the real-time graph structural features;

[0095] Real-time key features include real-time environmental impact features, real-time weather forecast impact features, real-time user behavior preference features, real-time distributed energy output features, real-time load demand features, real-time energy storage status features, and real-time microgrid interaction features;

[0096] S5-3: using the updated energy supply and demand forecasting module of the updated energy supply and demand forecasting model, performing energy supply and demand forecasting based on a number of real-time key features, and obtaining a real-time energy supply and demand forecasting result;

[0097] Real-time energy supply and demand forecast results include real-time energy supply forecast results (real-time power generation forecast of various distributed energy resources in the microgrid, such as solar energy, wind energy, and energy storage systems, and real-time power supply capacity forecast of the external power grid), real-time energy demand forecast results (real-time power demand forecast of industrial loads, commercial loads, and residential loads in the microgrid, and real-time power demand forecast of special loads such as electric vehicle charging stations and data centers), real-time energy storage status forecast results (real-time charging and discharging status forecast of the energy storage system, and real-time remaining capacity forecast of the energy storage system), real-time microgrid interaction forecast results (real-time interactive power forecast between the microgrid and the external power grid, including power purchase and sales, and real-time energy exchange forecast between subunits within the microgrid), and real-time distributed energy output forecast results (real-time output change trend and uncertainty forecast of various distributed energy resources).

[0098] S5-4: using the updated impact factor generation module of the updated energy management optimization model, analyzing the real-time energy supply and demand forecast results to obtain real-time impact factors, and setting real-time optimization targets based on the real-time impact factors;

[0099] In this embodiment, taking the obtained real-time influencing factors including energy supply loss and distributed energy output cost as an example, the real-time optimization goal is to minimize the energy supply loss and distributed energy output cost;

[0100] S5-5: using the updated fitness function updating module of the updated energy management optimization model, updating the preset fitness function of the updated energy management optimization model according to the real-time optimization target to obtain a real-time fitness function;

[0101] The formula of the real-time fitness function is:

[0102] Fit(P)=min[W1AX(P)+W2AC(P)]

[0103] Where Fit(P) is the real-time fitness function; AX(P) is the real-time energy supply loss function; AC(P) is the real-time distributed energy output cost function; P is the ISGA individual; W1, W2 are the first weight value and the second weight value;

[0104] S5-6: using the updated initial solution generation module of the updated energy management optimization model, transforming the real-time energy management optimization solution into individual vectors of the updated energy management optimization model, and generating initial solutions based on the individual vectors to obtain a plurality of initial solutions; the initial solutions correspond to the initial real-time energy management optimization solution;

[0105] The formula is:

[0106]

[0107] Where, P i The initial ISGA individual generated by the Circle chaotic map sequence, that is, the initial solution; P o is a randomly generated ISGA individual; i is the ISGA individual indicator; mod(*) is the remainder function;

[0108] S5-7: Using the updated iterative update module of the updated energy management optimization model, iteratively updating several initial solutions according to the real-time fitness function, and outputting the optimal solution with the best real-time fitness value, including the following steps:

[0109] S5-7-1: Use the real-time fitness function to obtain the initial real-time fitness value of each initial ISGA individual in the initial ISGA population, and take the initial ISGA individual with the lowest real-time fitness value as the leader goose;

[0110] S5-7-2: Entering the exploration phase, the leader goose rotation mechanism, the calling guidance mechanism, and the dynamic reverse mechanism are introduced to iteratively update the initial ISGA population to obtain an updated ISGA population and retain the optimal individual;

[0111] The leader goose rotation mechanism selects a new leader goose in each iteration based on the fitness value of the ISGA individuals. This mechanism can prevent the leader goose from falling into the local optimum too early and enhance the global search capability of the algorithm.

[0112] The formula is:

[0113]

[0114] Where, P i t+1Be the leader of an update; is the third-to-last initial ISGA individual in the initial ISGA population with the highest fitness value at the tth and t+1th iterations; is the fifth-to-last initial ISGA individual in the initial ISGA population with the highest fitness value at the tth iteration; t is the current iteration number; is the optimal individual; a is the first weight factor; rand is the random number generation function;

[0115] The calling guidance mechanism uses the sound wave propagation attenuation model to adjust the position update of the individual ISGA according to the distance between the individual and the leader goose. The position update of the ISGA individual that is closer is more affected by the leader goose, and it can quickly approach the optimal solution. The position update of the ISGA individual that is farther away is less affected by the leader goose, and it can maintain a certain exploration ability. This mechanism can avoid excessive aggregation or dispersion of the group and improve the local search accuracy of the algorithm.

[0116] The formula is:

[0117]

[0118] Where, It is an updated ISGA individual; is the initial ISGA individual of the tth iteration; is the sound intensity received by the initial ISGA individual; is the sound intensity parameter; L WA is the initial sound intensity; L low is the minimum acceptable sound intensity; a" is the convergence factor; is the initial ISGA individual with the farthest distance; r' is a random parameter; B(d) is the Brownian motion function; d is the Brownian motion parameter; is the XOR processing symbol;

[0119]

[0120] Where a" is the convergence factor; tanh(.) is the hyperbolic tangent function; t is the current number of iterations; t max is the maximum number of iterations; a max 、a min are the maximum and minimum values of the convergence factor, respectively; λ is the decreasing rate parameter, k' is the decreasing period parameter, λ = -2π, k' = π;

[0121] Dynamic reverse mechanism, which dynamically reverses the initial ISGA individuals to improve the diversity of exploration directions and avoid falling into local optimality;

[0122] The formula is:

[0123]

[0124] Where, is a reverse ISGA individual updated once; γ is the decreasing inertia coefficient; L max 、L min are the maximum and minimum values of the vector space respectively;

[0125] The leader goose of one update, several ISGA individuals of one update and several reverse ISGA individuals of one update are integrated to obtain an ISGA population of one update, and the ISGA individual with the lowest fitness value is retained as the optimal individual;

[0126] S5-7-3: Entering the development phase, introducing the abnormal boundary strategy and Gaussian mutation mechanism, performing a second update on the once-updated ISGA population, obtaining a second-updated ISGA population, and retaining the optimal individual;

[0127] Abnormal boundary strategy, calculates the difference between the fitness value of each updated ISGA individual and the average fitness value of the group. For ISGA individuals whose fitness value is much higher than the group average, their position update method will be adjusted, such as using Gaussian mutation mechanism, larger step size or smaller step size. This mechanism can help individuals avoid falling into local optimality and improve the convergence speed and accuracy of the algorithm;

[0128] The formula is:

[0129]

[0130] Where, is the second updated ISGA individual; is an updated ISGA individual; Fit(*) is the fitness function; Fit avg is the average fitness value of the group; is the ISGA individual with the highest fitness value; a' and e are the second and third weight factors; G(1,1) is the Gaussian mutation mechanism parameter;

[0131] S5-7-4: If the number of iterations is greater than or equal to the iteration threshold or the real-time fitness value of the optimal individual is less than the lower limit of the fitness threshold, the optimal individual is output as the optimal solution;

[0132] S5-8: Using the updated solution vector parsing module of the updated energy management optimization model, perform solution vector parsing on the individual vectors of the optimal solution to obtain the optimal real-time energy management optimization solution;

[0133] Real-time energy management optimization solutions include real-time energy allocation strategies (including real-time output adjustment instructions for each distributed energy source, real-time charging and discharging instructions for the energy storage system, and real-time interactive power adjustment between the microgrid and the external power grid), real-time load management strategies (real-time power demand response and adjustment for various loads, peak load management measures such as demand-side response and load shifting), real-time energy storage management strategies (real-time status monitoring and optimization control of the energy storage system, and real-time adjustment of the energy storage system's charging and discharging plan), real-time system stability control strategies (real-time monitoring and adjustment of system frequency and voltage, and real-time emergency response measures for emergencies and faults), and real-time energy efficiency optimization strategies (real-time optimization measures for energy utilization efficiency and real-time energy loss reduction strategies).

[0134] S6: Cloud data center, uses real-time topology network and dynamic data model to visualize real-time energy supply and demand forecast results and real-time energy management optimization plan, and synchronizes the real-time energy management optimization plan to all microgrid devices in the microgrid system through all edge computing gateways.

[0135] Example 2:

[0136] like Figure 2 As shown, this embodiment provides a microgrid system energy management optimization system based on big data analysis, which is used to implement a microgrid system energy management optimization method. The system includes a cloud data center, several edge computing gateways, several time series data acquisition devices and several microgrid devices. The cloud data center is connected to several edge computing gateways respectively, and each edge computing gateway is communicated with the time series data acquisition device and microgrid device of the corresponding monitoring point, and each time series data acquisition device is communicated with the corresponding microgrid device.

[0137] The cloud data center is used to perform big data analysis based on the collected multi-source time series big data using artificial intelligence algorithms, and to build a multi-source time series data processing model, energy supply and demand forecasting model, and energy management optimization model for the microgrid system; based on the alliance learning mechanism, the model is updated to obtain the updated multi-source time series data processing model, the updated energy supply and demand forecasting model, and the updated energy management optimization model, and the updated multi-source time series data processing model is synchronized to the edge computing gateway of all monitoring points; the real-time connection data of the microgrid system is collected, and the corresponding real-time topology network model is built according to the real-time connection data of the microgrid system, and all monitoring points are connected. The real-time standard time series data of the node is added to the real-time topology network model to obtain the real-time topology network and dynamic data model; based on the real-time topology network and dynamic data model, the updated energy supply and demand forecasting model is used to perform energy supply and demand forecasting; based on the obtained real-time energy supply and demand forecasting results, the updated energy management optimization model is used to generate an energy management optimization plan to obtain a real-time energy management optimization plan; the real-time topology network and dynamic data model are used to visualize the real-time energy supply and demand forecasting results and the real-time energy management optimization plan, and the real-time energy management optimization plan is synchronized to all microgrid devices in the microgrid system through all edge computing gateways;

[0138] The edge computing gateway is used to collect real-time multi-source time series data from monitoring points corresponding to the microgrid system, use the updated multi-source time series data processing model to process the real-time multi-source time series data, obtain real-time standard time series data, and upload the real-time standard time series data to the cloud data center;

[0139] The microgrid equipment in the microgrid system includes distributed energy, energy storage system, power load, energy management system, and grid-connected and off-grid switching devices;

[0140] Preferably, the cloud data center includes a model building unit, a model updating unit, a data adding unit, an energy supply and demand forecasting and energy management optimization unit, and a visualization unit, which are connected in sequence.

[0141] The model building unit is used to perform big data analysis based on the collected multi-source time series big data samples using artificial intelligence algorithms to build a multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model for the microgrid system;

[0142] A model updating unit is used to update the model according to the federated learning mechanism to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model, and synchronize the updated multi-source time series data processing model to the edge computing gateways of all monitoring points;

[0143] A data adding unit is used to collect real-time connection data of the microgrid system, build a corresponding real-time topology network model based on the real-time connection data of the microgrid system, and add the real-time standard time series data of all monitoring points to the real-time topology network model to obtain a real-time topology network and dynamic data model;

[0144] The energy supply and demand forecasting and energy management optimization unit is used to perform energy supply and demand forecasting based on the real-time topology network and dynamic data model using the updated energy supply and demand forecasting model, and to generate an energy management optimization plan based on the obtained real-time energy supply and demand forecast results using the updated energy management optimization model to obtain a real-time energy management optimization plan;

[0145] The visualization unit is used to visualize the real-time energy supply and demand forecast results and the real-time energy management optimization plan using the real-time topology network and dynamic data model, and synchronize the real-time energy management optimization plan to all microgrid devices in the microgrid system through all edge computing gateways.

[0146] The present invention provides a microgrid system energy management optimization method and system based on big data analysis. By utilizing big data analysis and artificial intelligence algorithms, it can efficiently process large-scale, multi-source, heterogeneous time series data, thereby improving the adaptability and response speed of the microgrid system to complex energy scenarios; with the help of the alliance learning mechanism, real-time updating of the model is achieved, ensuring that the prediction and optimization results can timely reflect the latest status of the microgrid system, and improving the accuracy of the results; through the real-time data processing and optimization scheme generation of the edge computing gateway, the real-time energy management of the microgrid system is achieved, meeting the needs of rapid response; the constructed energy supply and demand prediction model can discover deep features in time series data, improve the prediction accuracy, and perform better in complex and changeable energy supply and demand scenarios; the constructed energy management optimization model generates more flexible and effective energy management optimization schemes based on real-time energy supply and demand prediction results, which can adapt to real-time changes in energy supply and demand; by introducing edge computing gateways for distributed computing, the scalability and robustness of the system are enhanced, which can improve energy utilization efficiency, reduce operating costs and enhance system stability.

[0147] The present invention is not limited to the above optional embodiments. Anyone can derive various other forms of products based on the teachings of the present invention. The above specific embodiments should not be construed as limiting the scope of protection of the present invention. The scope of protection of the present invention shall be based on the scope defined in the claims, and the description can be used to interpret the claims.

Claims

1. A microgrid system energy management optimization method based on big data analysis, characterized by: The steps include: The cloud data center uses artificial intelligence algorithms to analyze the collected multi-source time series big data and build a multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model for the microgrid system. The cloud data center updates the model based on the federated learning mechanism to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model. It also synchronizes the updated multi-source time series data processing model to the edge computing gateways of all monitoring points. The edge computing gateway collects real-time multi-source time series data from monitoring points corresponding to the microgrid system, uses the updated multi-source time series data processing model to process the real-time multi-source time series data, obtains real-time standard time series data, and uploads the real-time standard time series data to the cloud data center; The cloud data center collects the real-time connection data of the microgrid system, builds the corresponding real-time topology network model based on the real-time connection data of the microgrid system, and adds the real-time standard time series data of all monitoring points to the real-time topology network model to obtain the real-time topology network and dynamic data model; The cloud data center uses the updated energy supply and demand forecasting model to perform energy supply and demand forecasting based on the real-time topology network and dynamic data model. Based on the obtained real-time energy supply and demand forecasting results, the updated energy management optimization model is used to generate an energy management optimization solution to obtain a real-time energy management optimization solution. The cloud data center uses real-time topology networks and dynamic data models to visualize real-time energy supply and demand forecast results and real-time energy management optimization solutions, and synchronizes the real-time energy management optimization solutions to all microgrid devices in the microgrid system through all edge computing gateways.

2. The microgrid system energy management optimization method based on big data analysis according to claim 1 is characterized by: The cloud data center uses artificial intelligence algorithms to analyze the collected multi-source time series big data and build a multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model for the microgrid system. The steps include: The cloud data center collects sample connection data of several microgrid systems and uses topology network simulation technology to build a corresponding sample topology network model based on the sample connection data. Collect sample multi-source time series big data including different data sources in the microgrid system, and preprocess them to obtain a number of preprocessed sample multi-source time series data; Based on several pre-processed sample multi-source time series data, a multi-source time series data processing model is constructed using deep learning and reinforcement learning algorithms, and several sample standard time series data are obtained; Based on several sample topological network models and several sample standard time series data, a deep learning and feature screening algorithm is used to build an energy supply and demand forecasting model and obtain several sample energy supply and demand forecasting results; Based on the energy supply and demand forecast results of several samples, an energy management optimization model is constructed using deep learning and swarm intelligence optimization algorithms.

3. The microgrid system energy management optimization method based on big data analysis according to claim 2 is characterized by: The multi-source time series data processing model is constructed based on the CNN-LSTM-MOGRPO-DPA algorithm, and the multi-source time series data processing model includes an unstructured data feature extraction module constructed based on the CNN algorithm, a structured data feature extraction module constructed based on the LSTM algorithm, a data processing strategy generation module constructed based on the MOGRPO algorithm, and a data processing algorithm library constructed based on the DPA algorithm. The data processing algorithm library includes several DPA algorithm packages.

4. The microgrid system energy management optimization method based on big data analysis according to claim 3 is characterized by: The energy supply and demand forecasting model is constructed based on the GCN-RF-MLP algorithm, and the energy supply and demand forecasting model includes a graph structure feature extraction module constructed based on the GCN algorithm, a key feature screening module constructed based on the RF algorithm, and an energy supply and demand forecasting module constructed based on the MLP algorithm, which are connected in sequence.

5. The microgrid system energy management optimization method based on big data analysis according to claim 4 is characterized in that: The energy management optimization model is constructed based on the ISGA algorithm, and the energy management optimization model includes an influence factor generation module, a fitness function update module, an initial solution generation module, an iterative update module and a solution vector analysis module connected in sequence.

6. The microgrid system energy management optimization method based on big data analysis according to claim 5, characterized in that: The cloud data center updates the model based on the federated learning mechanism to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model. The updated multi-source time series data processing model is then synchronized to the edge computing gateways at all monitoring points, including the following steps: The cloud data center builds a corresponding historical topology network model based on the historical connection data of the microgrid system, extracts the initial model parameters of the historical topology network model, multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model, and sends them to all edge computing gateways; The edge computing gateway reconstructs and deploys the initial model parameters of the historical topology network model, multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model based on the initial model parameters. The edge computing gateway optimizes and trains the multi-source time series data processing model, energy supply and demand prediction model, and energy management optimization model based on the historical topology network model and several historical multi-source time series data collected. It extracts the optimized training model parameters after optimization training and uploads them to the cloud data center. The cloud data center integrates the optimized training model parameters uploaded by all edge computing gateways to obtain comprehensive model parameters. Based on the comprehensive model parameters, the cloud data center updates the model to obtain an updated multi-source time series data processing model, an updated energy supply and demand forecasting model, and an updated energy management optimization model. The cloud data center extracts the updated model parameters of the updated multi-source time series data processing model and sends the updated model parameters to the edge computing gateways of all monitoring points to complete the synchronization of the updated multi-source time series data processing model.

7. The microgrid system energy management optimization method based on big data analysis according to claim 6 is characterized by: The edge computing gateway collects real-time multi-source time series data from the monitoring points corresponding to the microgrid system, uses the updated multi-source time series data processing model to process the real-time multi-source time series data to obtain real-time standard time series data, and uploads the real-time standard time series data to the cloud data center, including the following steps: The edge computing gateway collects real-time multi-source time series data from monitoring points corresponding to the microgrid system and inputs the real-time multi-source time series data into the updated multi-source time series data processing model; Using the updated unstructured data feature extraction module of the updated multi-source time series data processing model to extract real-time unstructured data features of the real-time multi-source time series data; Using the updated structured data feature extraction module of the updated multi-source time series data processing model to extract real-time structured data features of the real-time multi-source time series data; Based on the real-time unstructured data characteristics and real-time structured data characteristics, the data processing strategy generation module of the updated multi-source time series data processing model is used to generate a real-time data processing strategy; According to the real-time data processing strategy, several target DPA algorithm packages are called in the data processing algorithm library, and according to several target DPA algorithm packages, real-time multi-source time series data are processed to obtain real-time standard time series data, and the real-time standard time series data is uploaded to the cloud data center.

8. The microgrid system energy management optimization method based on big data analysis according to claim 7 is characterized in that: The cloud data center uses the updated energy supply and demand forecasting model to perform energy supply and demand forecasting based on the real-time topology network and dynamic data model. Based on the obtained real-time energy supply and demand forecast results, the updated energy management optimization model is used to generate an energy management optimization solution. The obtained real-time energy management optimization solution includes the following steps: Cloud data center,The cloud data center uses the updated graph structure feature extraction module of the updated energy supply and demand prediction model to extract the real-time graph structure features of the real-time topology network and dynamic data model; Extracting several real-time key features of the real-time graph structural features using the updated key feature screening module of the updated energy supply and demand forecasting model; An updated energy supply and demand forecasting module of the updated energy supply and demand forecasting model is used to forecast energy supply and demand based on a number of real-time key features to obtain a real-time energy supply and demand forecasting result; Using the updated impact factor generation module of the updated energy management optimization model, the real-time energy supply and demand forecast results are analyzed to obtain the real-time impact factors, and the real-time optimization goals are set based on the real-time impact factors; Using an updated fitness function updating module of the updated energy management optimization model, the preset fitness function of the updated energy management optimization model is updated according to the real-time optimization target to obtain a real-time fitness function; Using an updated initial solution generation module of the updated energy management optimization model, the real-time energy management optimization solution is converted into individual vectors of the updated energy management optimization model, and initial solutions are generated based on the individual vectors to obtain a plurality of initial solutions; the initial solutions correspond to the initial real-time energy management optimization solution; Iteratively updating a number of initial solutions according to a real-time fitness function using an updated iterative update module of the updated energy management optimization model, and outputting an optimal solution with the best real-time fitness value; The updated solution vector parsing module of the updated energy management optimization model is used to perform solution vector parsing on the individual vectors of the optimal solution to obtain the optimal real-time energy management optimization solution.

9. A microgrid system energy management optimization system based on big data analysis, used to implement the microgrid system energy management optimization method according to any one of claims 1 to 8, characterized in that: The system includes a cloud data center, several edge computing gateways, several time series data acquisition devices and several microgrid devices. The cloud data center is connected to several edge computing gateways respectively, each of the edge computing gateways is communicatively connected to the time series data acquisition device and microgrid device of the corresponding monitoring point, and each of the time series data acquisition devices is communicatively connected to the corresponding microgrid device.

10. The microgrid system energy management optimization system based on big data analysis according to claim 9, characterized in that: The cloud data center includes a model building unit, a model updating unit, a data adding unit, an energy supply and demand prediction and energy management optimization unit, and a visualization unit which are connected in sequence.

Images