A power system planning and operation optimization method, system, device and storage medium

Abstract

The present application relates to the technical field of power energy storage system planning and analysis, and discloses a power system planning operation optimization method, system, device and storage medium. The method includes: collecting electrical, equipment status and environmental parameters for feature extraction to evaluate the health status of the equipment; grouping distributed energy based on the health status evaluation and weather warning information to form a virtual power plant resource pool; decomposing tasks according to the characteristics of the resource pool to formulate a collaborative scheduling strategy; arranging a maintenance plan according to the equipment status and the scheduling strategy; optimizing the system operation parameters based on the scheduling strategy and the maintenance plan to obtain a balance control scheme. The present application realizes the dynamic balance of minimizing the operation cost of the distribution network and maximizing the consumption of renewable energy through the intelligent orchestration technology of the virtual power plant group while considering the equipment health status and the impact of extreme weather.

Description

Technical Field

[0001] This application relates to the technical field of power energy storage system planning and analysis, and particularly to a power system planning operation optimization method, system, device, and storage medium. Background Art

[0002] The planning operation optimization of the power system is an important link to ensure the safe and stable operation of the power grid. The existing power system planning operation methods mainly carry out optimization from aspects such as equipment maintenance, load forecasting, and power source scheduling, and optimize and calculate the system operation mode by establishing a mathematical model. At the same time, with the large-scale access of distributed energy, the virtual power plant technology is gradually applied to the coordinated control of the power system, and the regulation ability of the system is improved by integrating scattered distributed resources.

[0003] However, the following deficiencies exist in the existing technology: there is a lack of an effective coordination mechanism between equipment health status assessment and scheduling optimization, resulting in difficulty in coordinating maintenance plans with system operation requirements; secondly, there is a lack of quantitative assessment means for the impact of extreme weather on power equipment, making it difficult to formulate targeted prevention and control strategies; thirdly, the resource grouping method of the virtual power plant is relatively simple, and factors such as equipment health status and weather impact are not fully considered, reducing the operation reliability of the system; finally, the multi-time scale coordinated optimization mechanism is imperfect, making it difficult to achieve the global optimization of system operation. Summary of the Invention

[0004] This application provides a power system planning operation optimization method, system, device, and storage medium, which is used to achieve the dynamic balance of minimizing the operation cost of the distribution network and maximizing the consumption of renewable energy through the intelligent orchestration technology of the virtual power plant group while considering the equipment health status and the impact of extreme weather.

[0005] In a first aspect, this application provides a power system planning operation optimization method, and the power system planning operation optimization method includes: multi-dimensionally collecting and extracting features from power system operation data according to electrical parameters, equipment status parameters, and environmental parameters to obtain equipment health status assessment data; performing resource grouping and priority division on distributed energy, controllable loads, and energy storage systems based on the equipment health status assessment data and regional extreme weather warning information to form a virtual power plant resource pool; performing global task decomposition and target allocation according to the capacity distribution and operation characteristics of the virtual power plant resource pool to generate a virtual power plant group collaborative scheduling strategy; predicting the status of equipment and arranging maintenance time sequences based on the equipment health status assessment data and the virtual power plant group collaborative scheduling strategy to obtain a critical equipment maintenance plan; and performing multi-time scale optimization configuration on system operation parameters based on the virtual power plant group collaborative scheduling strategy and the critical equipment maintenance plan to obtain a power system balance control plan.

[0006] In a second aspect, the present application provides a power system planning and operation optimization system, and the power system planning and operation optimization system includes:

[0007] An extraction module, configured to perform multi-dimensional acquisition and feature extraction on power system operation data according to electrical parameters, equipment status parameters, and environmental parameters, so as to obtain equipment health status evaluation data;

[0008] A partitioning module, configured to perform resource grouping and priority partitioning on distributed energy, controllable loads, and energy storage systems based on the equipment health status evaluation data and regional extreme weather warning information, so as to form a virtual power plant resource pool;

[0009] An allocation module, configured to perform global task decomposition and target allocation according to the capacity distribution and operation characteristics of the virtual power plant resource pool, so as to generate a cooperative scheduling strategy for a virtual power plant group;

[0010] A prediction module, configured to perform status prediction and maintenance timing arrangement on equipment according to the equipment health status evaluation data and the cooperative scheduling strategy of the virtual power plant group, so as to obtain a maintenance plan for key equipment;

[0011] A configuration module, configured to perform multi-time-scale optimization configuration on system operation parameters based on the cooperative scheduling strategy of the virtual power plant group and the maintenance plan for key equipment, so as to obtain a power system balance control scheme.

[0012] In a third aspect of the present invention, there is provided a computer device, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory, so that the computer device executes the above-mentioned power system planning and operation optimization method.

[0013] In a fourth aspect of the present invention, there is provided a computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium, and when the instructions run on a computer, the computer is enabled to execute the above-mentioned power system planning and operation optimization method.

[0014] In the technical solution provided by this application, through multi-dimensional collection and feature extraction of electrical parameters, equipment status parameters, and environmental parameters, a comprehensive equipment health status evaluation system is established, improving the accuracy and timeliness of equipment status monitoring; based on the equipment health status evaluation data and regional extreme weather warning information, a method of resource grouping and priority division is used to achieve dynamic optimization allocation of virtual power plant resources, enhancing the system's adaptability to extreme weather; through global task decomposition and target allocation of the capacity distribution and operation characteristics of the virtual power plant resource pool, a scheduling strategy generation method based on multi-level collaboration is proposed, improving the coordination control effect of the system; according to the equipment health status evaluation data and the virtual power plant group collaborative scheduling strategy, state prediction and maintenance timing arrangement of equipment are carried out, constructing a preventive maintenance decision-making mechanism, reducing the equipment failure risk; based on the virtual power plant group collaborative scheduling strategy and the critical equipment maintenance plan, multi-time scale optimization allocation of system operation parameters is carried out, designing a dynamic balance control scheme, and achieving the improvement of system operation efficiency. In this invention, a deep learning feature fusion model is adopted in data processing. This model can effectively extract the key features of the equipment operation state, dynamically weight different features through the attention mechanism, and improve the accuracy of state evaluation; in resource scheduling, a graph neural network algorithm is applied for dynamic grouping. This algorithm fully considers the topological relationship and operation characteristics among resources, enhancing the collaborative control effect of the virtual power plant; in optimization decision-making, a multi-agent reinforcement learning method is introduced to achieve adaptive optimization control of the system, enhancing the practicability and popularization of the solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0016] Figure 1 It is a schematic diagram of an embodiment of the power system planning operation optimization method in the embodiments of this application;

[0017] Figure 2 It is a schematic flowchart of multi-time scale optimization allocation of system operation parameters in the embodiments of this application;

[0018] Figure 3 It is a schematic diagram of an embodiment of the power system planning operation optimization system in the embodiments of this application;

[0019] Figure 4 It is a schematic block diagram of the structure of a computer device in the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] The embodiments of the present application provide a method, a system, a device, and a storage medium for optimizing the planning and operation of a power system. Terms such as "first", "second", "third", "fourth", etc. (if any) in the specification, claims, and the above-mentioned drawings of the present application are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" or "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily limit to those clearly listed steps or units, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0021] For ease of understanding, the specific process of the embodiments of the present application is described below. Please refer to Figure 1 , an embodiment of the method for optimizing the planning and operation of a power system in the embodiments of the present application includes:

[0022] Step S101: Multidimensionally collect and extract features from the power system operation data according to electrical parameters, equipment status parameters, and environmental parameters to obtain equipment health status evaluation data;

[0023] Step S102: Group resources and divide priorities for distributed energy, controllable loads, and energy storage systems based on the equipment health status evaluation data and regional extreme weather warning information to form a virtual power plant resource pool;

[0024] Step S103: Decompose global tasks and allocate targets according to the capacity distribution and operation characteristics of the virtual power plant resource pool to generate a collaborative scheduling strategy for the virtual power plant group;

[0025] Step S104: Predict the status of equipment and arrange maintenance time sequences based on the equipment health status evaluation data and the collaborative scheduling strategy of the virtual power plant group to obtain a maintenance plan for key equipment;

[0026] Step S105: Optimally configure the system operation parameters on multiple time scales based on the collaborative scheduling strategy of the virtual power plant group and the maintenance plan for key equipment to obtain a power system balance control scheme.

[0027] It can be understood that the execution subject of the present application can be a system for optimizing the planning and operation of a power system, or a terminal or a server. Specifically, it is not limited here. The embodiments of the present application are described by taking the server as the execution subject as an example.

[0028] Specifically, multi-dimensional data collection and feature extraction are carried out. Electrical parameters include voltage, current, active power, and reactive power. Equipment status parameters include temperature, vibration, noise, and partial discharge data. Environmental parameters include air temperature, humidity, wind speed, and rainfall data. Standardize the electrical parameter data, normalize parameters such as voltage and current to a unified dimension, and form an electrical feature vector. Conduct time-frequency domain analysis on the equipment status parameters, extract the spectral features and time-domain statistical features of the vibration signal through wavelet transform, normalize data such as temperature and noise, and construct an equipment status feature vector. Normalize the environmental parameter data to form an environmental feature vector. Integrate the three types of feature vectors to construct an equipment operation feature matrix. Conduct time-series correlation analysis on the dynamic data in the feature matrix to obtain a dynamic health index, conduct a life cycle assessment on the static data to obtain a static health index, and obtain equipment health status assessment data through weighted integration. Based on the equipment health status assessment data, extract the health risk coefficient for each equipment to form an equipment vulnerability assessment matrix. At the same time, identify and classify the types and intensity levels of weather disasters according to the regional extreme weather warning information to obtain a weather impact factor set. Conduct correlation analysis between the equipment vulnerability assessment matrix and the weather impact factor set to construct an equipment-weather sensitivity table, which reflects the affected degree of different types of equipment under various extreme weather conditions. Conduct regional clustering on distributed energy, controllable loads, and energy storage systems according to the equipment-weather sensitivity table to obtain a preliminary resource grouping. Evaluate the available capacity and analyze the response capabilities of various resources to generate resource regulation priority indicators. Finally, integrate and map the resource grouping and priority indicators to form a virtual power plant resource pool.

[0029] Based on the virtual power plant resource pool, conduct statistical analysis on the capacity distribution data to form a resource capacity basis matrix, and conduct volatility and stability assessments to obtain a capacity availability assessment table. Analyze the operation characteristics of various resources, such as output characteristics, response time, and regulation capabilities, and construct operation characteristic evaluation indicators. Conduct multi-objective correlation analysis between the capacity availability assessment table and the operation characteristic evaluation indicators to obtain a resource-task fitness matrix. Conduct hierarchical decomposition and boundary identification on this matrix to generate a task partition table and a cross-region mutual assistance plan. Based on these results, conduct resource allocation and load distribution to obtain an initial scheduling instruction set. Finally, generate a collaborative scheduling strategy for the virtual power plant group through the collaborative optimization of time series and spatial regions.

[0030] Based on the equipment health status assessment data, analyze the historical trends, extract fault characteristics, and generate the equipment health degradation curve. Correlate this curve with the equipment scheduling load in the collaborative scheduling strategy of the virtual power plant group to obtain the equipment operation pressure index. Based on this index, conduct remaining life prediction and risk rating, and construct the equipment status warning table. Classify and sort the equipment according to the importance and operation risk to generate the maintenance priority matrix. Combine the collaborative scheduling strategy of the virtual power plant group to divide the maintenance time window and obtain the maintenance time sequence arrangement table. Finally, conduct maintenance resource allocation and process planning to obtain the maintenance plan for key equipment. Decompose the collaborative scheduling strategy of the virtual power plant group according to different time scales to obtain the hierarchical scheduling strategy matrix. Conduct time series analysis on the scheduling data of each time scale in this matrix to generate the scheduling strategy time series characteristic table. Map the maintenance plan for key equipment in the time dimension to generate the equipment available capacity time series table, and conduct capacity balance analysis according to time nodes to obtain the equipment operation availability curve. Calculate the resource matching degree according to the scheduling strategy time series characteristic table and the equipment operation availability curve, and construct the resource allocation constraint matrix. Conduct multi-dimensional constraint analysis on this matrix to form the system operation boundary condition set. Conduct regional coordination evaluation of the spatial distribution of the boundary condition set to obtain the regional mutual assistance constraint table. Finally, conduct boundary condition integration and constraint mapping to generate the system operation constraint boundary set, and based on this, conduct multi-time scale optimal allocation to obtain the power system balance control plan.

[0031] For example, when planning and optimizing the operation of a regional distribution network, collect multi-dimensional data such as the voltage, current, temperature, and vibration of the main transformer. Through data standardization and feature extraction, it is found that the harmonic content of the A-phase current is relatively high and the temperature fluctuates significantly. Combining with the weather forecast showing that there is strong convective weather in this area, through equipment-weather sensitivity analysis, it is shown that the fault risk of this transformer is relatively high under strong convective weather. Therefore, when grouping resources, prioritize the distributed photovoltaic, energy storage system, and controllable air-conditioning load in the area where this transformer is located to form a virtual power plant unit. Through the analysis of the capacity characteristics and response capabilities of this unit, a time-of-use scheduling strategy is formulated. At the same time, according to the equipment health status assessment results, list this transformer in the priority sequence of the maintenance plan and select to conduct maintenance during the low valley period of photovoltaic power generation. Finally, through multi-time scale optimal allocation, a system balance control plan including maintenance arrangements, load regulation, energy storage scheduling, etc. is formed.

[0032] In the embodiments of the present application, by performing multi-dimensional acquisition and feature extraction on electrical parameters, equipment status parameters, and environmental parameters, an equipment health status evaluation system is established, improving the accuracy and timeliness of equipment status monitoring; based on the equipment health status evaluation data and regional extreme weather warning information, a method of resource grouping and priority division is carried out, realizing the dynamic optimal allocation of virtual power plant resources and enhancing the system's adaptability to extreme weather; by performing global task decomposition and target allocation on the capacity distribution and operation characteristics of the virtual power plant resource pool, a scheduling strategy generation method based on multi-level collaboration is proposed, improving the coordinated control effect of the system; according to the equipment health status evaluation data and the virtual power plant group collaborative scheduling strategy, state prediction and maintenance time sequence arrangement of the equipment are carried out, constructing a preventive maintenance decision-making mechanism and reducing the equipment failure risk; based on the virtual power plant group collaborative scheduling strategy and the critical equipment maintenance plan, multi-time scale optimal configuration of the system operation parameters is carried out, designing a dynamic balance control scheme and realizing the improvement of the system operation efficiency. In the data processing aspect of the present invention, a deep learning feature fusion model is adopted, which can effectively extract the key features of the equipment operation state, dynamically weight different features through the attention mechanism, and improve the accuracy of state evaluation; in the resource scheduling aspect, a graph neural network algorithm is applied for dynamic grouping, which fully considers the topological relationship and operation characteristics among resources and improves the collaborative control effect of the virtual power plant; in the optimization decision-making aspect, a multi-agent reinforcement learning method is introduced to realize the adaptive optimal control of the system and enhance the practicability and popularization of the scheme.

[0033] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0034] (1) Collect the voltage, current, active power, and reactive power data in the electrical parameters, perform standardization processing on the collected data, and obtain the electrical feature vector;

[0035] (2) Collect the equipment temperature, vibration, noise, and partial discharge data in the equipment status parameters, perform time-frequency domain analysis on the collected data through wavelet transform, and obtain the equipment status feature vector;

[0036] (3) Collect the air temperature, humidity, wind speed, and rainfall data in the environmental parameters, perform normalization processing on the collected data, and obtain the environmental feature vector;

[0037] (4) Perform feature fusion on the electrical feature vector, equipment status feature vector, and environmental feature vector to construct an equipment operation feature matrix;

[0038] (5) Perform time series correlation analysis on the dynamic data in the equipment operation feature matrix to obtain the equipment dynamic health index;

[0039] (6) Conduct a life cycle assessment on the static data in the device operation characteristic matrix to obtain the device static health index;

[0040] (7) Perform weighted fusion of the device dynamic health index and the device static health index to obtain the device health status assessment data.

[0041] Specifically, during the electrical parameter acquisition process, voltage, current, active power, and reactive power data are obtained in real time from each node of the substation, transmission line, and distribution network through the intelligent sensor network. These raw data have problems such as inconsistent dimensions and large differences in numerical ranges, so standardization processing is required. The Z-score standardization method is used for standardization processing, converting each parameter into a standard distribution with a mean of 0 and a standard deviation of 1, enabling effective comparison and fusion of electrical parameters with different dimensions. For example, the voltage data of a certain transformer fluctuates between 110 kV and 115 kV, and the current data changes between 90 A and 110 A. After standardization processing, these data are mapped to the same numerical interval to form an electrical feature vector. In the device status parameter acquisition stage, the device temperature, vibration, noise, and partial discharge data are mainly monitored. The device temperature is obtained through an infrared thermal imager and a temperature sensor, the vibration data is measured by an acceleration sensor, the noise data is collected through an acoustic sensor, and the partial discharge data is obtained by a special discharge detection instrument. These device status data often contain rich time-frequency characteristics, and key features of the device operation status can be extracted through time-frequency domain analysis using wavelet transform. The wavelet transform decomposes the signal into wavelet coefficients of different frequencies and time scales, and can effectively capture the non-stationary characteristics and transient features in the device status data. By selecting appropriate wavelet basis functions, such as Daubechies wavelet or Morlet wavelet, multi-scale decomposition is performed on the vibration signal to extract key features such as energy distribution, spectral characteristics, and singular points, forming a device status feature vector. In the environmental parameter acquisition link, the air temperature, humidity, wind speed, and rainfall data are mainly concerned, and these parameters directly affect the device operation environment. The environmental data is obtained by weather stations and environmental monitoring devices distributed at key nodes in the power system. Due to the large differences in the numerical ranges of environmental parameters, normalization processing is required to map each parameter value to the interval [0, 1] for subsequent fusion analysis. The maximum-minimum normalization method is used for normalization processing to calculate the normalized value of each environmental parameter, forming an environmental feature vector.

[0042] Fuse the electrical feature vector, device status feature vector, and environmental feature vector to construct a device operation feature matrix. The feature fusion adopts a multi-level fusion strategy. First, weight configuration is performed on each feature vector, and weight coefficients are assigned according to the influence degree of each parameter on the device health status. Then, the three types of feature vectors are fused into a unified feature representation through weighted summation. The device operation feature matrix contains both the dynamic features of the current operation status of the device and the static features of the historical operation and basic attributes of the device. Perform time series correlation analysis on the dynamic data in the device operation feature matrix, use the autoregressive moving average model to process the time series data, extract the trend, periodicity, and randomness features of the time series, calculate the time series correlation coefficient matrix between parameters, and identify the key parameter combinations and their change patterns. Determine whether there is an abnormality in the device operation status by setting a threshold, and combine the expert knowledge base to evaluate the risk level of the abnormality, and finally obtain the device dynamic health index. Perform life cycle assessment on the static data in the device operation feature matrix. Based on the static information such as the model, operation years, cumulative operation time, and historical maintenance records of the device, combine the device aging curve model to calculate the theoretical life consumption rate of the device. At the same time, consider factors such as the importance of the device, spare part supply situation, and maintenance difficulty, and comprehensively evaluate to obtain the device static health index.

[0043] Fuse the device dynamic health index and the device static health index through weighting to obtain comprehensive device health status assessment data. The weighting fusion process takes into account the characteristics of different types of devices, and assigns different weights to the dynamic index and the static index. For example, for large static devices such as transformers, the weight of the static health index may be higher, while for devices such as circuit breakers that are frequently operated, the weight of the dynamic health index may be greater.

[0044] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0045] (1) Analyze the device health status assessment data, extract the health risk coefficient of each device, and generate a device vulnerability assessment matrix;

[0046] (2) Identify the type of weather disaster and divide the intensity level according to the regional extreme weather warning information to obtain a set of weather impact factors, and perform correlation analysis between the device vulnerability assessment matrix and the set of weather impact factors to construct a device-weather sensitivity table;

[0047] (3) Based on the device-weather sensitivity table, perform regional clustering on distributed energy, controllable loads, and energy storage systems to obtain a resource grouping sequence, and perform available capacity assessment and response ability analysis on various resources in the resource grouping sequence to generate a resource regulation priority index;

[0048] (4) Conduct geographical location correlation analysis on each group of resources in the resource grouping sequence to generate a resource spatial distribution matrix;

[0049] (5) Cross-map the resource spatial distribution matrix with the resource regulation priority index to obtain an initial resource combination table;

[0050] (6) Conduct energy complementarity analysis on the resources in the initial resource combination table, construct a resource complementarity degree matrix, and optimize and reorganize the initial resource combination table according to the resource complementarity degree matrix to obtain a resource synergy index;

[0051] (7) Perform weighted fusion of the resource synergy index and the resource regulation priority index to form a virtual power plant resource pool.

[0052] Specifically, conduct risk analysis on the equipment health status data. For each piece of equipment, calculate the risk coefficient based on its health status assessment data:

[0053] ；

[0054] where, is the equipment risk coefficient (dimensionless, range 0-1); is the weight of the i-th health status indicator; is the deviation value of this indicator; is the number of evaluation indicators. These risk coefficients form an equipment vulnerability assessment matrix.

[0055] Analyze the extreme weather warning information and construct a weather impact factor:

[0056] ;

[0057] where, is the weather impact factor; is the type coefficient of the j-th type of weather disaster; is the intensity level coefficient; is the duration impact index; is the number of weather disaster types. Correlate the weather impact factor with the equipment vulnerability to obtain a sensitivity calculation formula:

[0058] ;

[0059] where, is the sensitivity of the i-th equipment to the j-th type of weather; is the coupling coefficient. This constitutes an equipment-weather sensitivity table.

[0060] Conduct resource clustering analysis based on the sensitivity table:

[0061] ;

[0062] Among them, is the eigenvalue of the th class of resource group; is the resource capacity; is the quantity of this class of resources. Evaluate the available capacity and response characteristics of each class of resources:

[0063] ;

[0064] Among them, is the resource regulation priority index; is the available capacity coefficient; is the response time coefficient; is the regulation error coefficient.

[0065] Conduct geographical location correlation analysis and construct a spatial distribution matrix:

[0066] ;

[0067] Among them, is the spatial distance (km) between resources; ( , ) and ( , ) are the longitude and latitude coordinates of two resource points.

[0068] Conduct energy complementarity analysis on resources:

[0069] ;

[0070] Among them, is the complementarity coefficient; and are the outputs of two classes of resources at time t; T is the length of the evaluation period.

[0071] The final resource pool construction adopts weighted fusion:

[0072] ；

[0073] Among them, is the comprehensive evaluation index of the resource pool; and are the weight coefficients and + = 1. This indicator comprehensively considers the regulation priority and complementary characteristics of resources, providing a basis for the optimal allocation of resources in a virtual power plant. Taking a photovoltaic power station, a wind farm, and a energy storage power station in a certain area as an example, the sensitivity of various resources to extreme weather can be identified through the above evaluation system, and the resource combination can be reasonably configured to improve the operation reliability and economy of the virtual power plant. This evaluation system systematically integrates multi-dimensional factors such as equipment status, weather impact, geographical distribution, and energy characteristics to form a method for constructing a resource pool.

[0074] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0075] (1) Hierarchically statistically analyze the capacity distribution data in the virtual power plant resource pool according to energy type and capacity size, extract spatio-temporal distribution characteristics, and generate a resource capacity basic matrix;

[0076] (2) Analyze the volatility and evaluate the stability of the data in the resource capacity basic matrix, and compare it with historical operation data to obtain a capacity availability evaluation table;

[0077] (3) According to the output characteristics, response time, and regulation ability of various resources in the virtual power plant resource pool, extract multi-dimensional parameters and perform feature mapping to construct an operation characteristic evaluation index;

[0078] (4) Conduct a multi-objective correlation analysis of the capacity availability evaluation table and the operation characteristic evaluation index to obtain a resource-task fitness matrix;

[0079] (5) Perform hierarchical decomposition and boundary recognition on the resource-task fitness matrix to generate a task partition table and an over-area mutual assistance plan;

[0080] (6) Based on the task partition table and the over-area mutual assistance plan, conduct resource allocation and load distribution to obtain an initial scheduling instruction set;

[0081] (7) Coordinate and optimize the initial scheduling instruction set according to time series and spatial regions, and resolve conflicts to generate a collaborative scheduling strategy for the virtual power plant group.

[0082] Specifically, when hierarchically statistical analyzing the capacity distribution data in the virtual power plant resource pool, various distributed energy sources, controllable loads, and energy storage devices need to be classified and summarized according to energy types (such as photovoltaic, wind power, energy storage, adjustable load, etc.) and capacity sizes (such as large, medium, and small). The hierarchical statistics adopt a tree structure, and the resources are classified at multiple levels according to the primary energy type, secondary capacity level, and tertiary geographical location to form structured data. The extraction of spatio-temporal distribution characteristics is to statistically analyze the geographical location and time availability of various resources, including the density distribution of resources in different regions, seasonal change characteristics, and intraday fluctuation rules, etc. By matrix processing these characteristic data, a resource capacity base matrix is formed. The rows of this matrix represent different resource types, the columns represent different spatio-temporal characteristic dimensions, and the matrix element values represent the available capacity values of the corresponding type of resource under specific spatio-temporal conditions. When analyzing the volatility of the data in the resource capacity base matrix, the coefficient of variation method is used to calculate the volatility index of the capacity of various resources. The larger the coefficient of variation, the stronger the output volatility of this type of resource. At the same time, statistical methods such as standard deviation analysis and entropy method are used to evaluate the stability of the resource capacity, and the stability scores of various resources are obtained. The results of volatility analysis and stability assessment are compared with historical operation data. The sliding time window method is used to compare the differences in the current resource state with that in the same historical period and under the same weather conditions, and the similarity index is calculated. Based on these analysis results, a capacity availability assessment table is generated. This table includes the actual available capacity ratio and reliability level of various resources in different time periods and under different weather conditions.

[0083] When extracting multi-dimensional parameters according to the output characteristics, response time, and regulation ability of various resources in the virtual power plant resource pool, it is necessary to model and quantify the dynamic response characteristics of each type of resource. The output characteristics include parameters such as maximum output, minimum output, ramp rate, and duration; the response time refers to the time delay from receiving the dispatch instruction to the actual response; the regulation ability reflects the execution accuracy and stability of the resource to the dispatch instruction. By normalizing these parameters and assigning weights, and using multi-dimensional feature mapping methods such as radar charts, an evaluation index system for resource operation characteristics is constructed to form the ability portraits of various resources in different operation scenarios. When conducting multi-objective correlation analysis between the capacity availability assessment table and the operation characteristic evaluation index, the fuzzy comprehensive evaluation method is used to cross-match the two sets of index systems. First, a task type library is set up, including different task types such as peak shaving, frequency modulation, standby, and demand response. Then, according to the specific demand characteristics of each task type, the matching degree with the operation characteristics of the resources is calculated to generate a resource-task matching degree matrix. The rows of this matrix represent different resources, the columns represent different task types, and the matrix element values represent the matching degree of a specific resource to execute a specific task.

[0084] When performing hierarchical decomposition on the resource-task fitness matrix, the clustering analysis method is used to group resources and tasks with similar fitness characteristics to form resource-task sub-blocks. Boundary recognition is achieved by setting a fitness threshold to identify the boundary resources and tasks of each sub-block, clarifying the task boundaries of each resource group and the resource boundaries of each task type. The results of hierarchical decomposition and boundary recognition are used to generate a task partition table, clarifying the task allocation plan for different regions and different time periods. The cross-region mutual assistance plan is based on the boundary recognition results, conducts complementary analysis on the resources and tasks in the boundary region, and formulates a cross-region resource allocation strategy in extreme situations. When performing resource allocation and load distribution based on the task partition table and the cross-region mutual assistance plan, the linear programming method is used to solve the optimal resource allocation plan. With the objective function of minimizing the scheduling cost and maximizing the renewable energy consumption rate, and subject to power balance constraints, line capacity constraints, and equipment operation constraints, the optimal output plan of various resources and the response plan of various loads are calculated to form an initial scheduling instruction set.

[0085] When co-optimizing and resolving conflicts for the initial scheduling instruction set according to the time series and spatial regions, the rolling time-domain decomposition method is used to handle the instruction continuity in the time dimension to ensure the smooth transition of scheduling instructions in each time period; the regional coordination mechanism is used to handle the resource allocation conflicts in the spatial dimension to ensure the power balance and safety constraints across regions. Through iterative optimization, the spatio-temporal conflicts between scheduling instructions are eliminated, and finally a collaborative scheduling strategy for the virtual power plant group is generated.

[0086] Taking a certain regional power grid as an example, this region includes a photovoltaic power station (total capacity of 50 MW), a wind farm (total capacity of 30 MW), industrial controllable loads (total capacity of 20 MW), and an energy storage system (capacity of 15 MW / 60 MWh). First, these resources are hierarchically counted by type and capacity, and combined with geographical distribution and time characteristics to form a resource capacity base matrix. By analyzing the historical output data of photovoltaic and wind power, it is calculated that the capacity variation coefficient of the photovoltaic power station is 0.3 on sunny days and rises to 0.6 under cloudy weather; the variation coefficient of the wind farm is 0.2 under stable meteorological conditions and rises to 0.5 when the weather changes violently. Combining weather forecast data and historical performance in the same period, a capacity availability assessment table for each time period of the next day is generated. Analyze the operating characteristics of various resources, such as short response time of photovoltaic but unable to increase output, certain regulation margin of wind power, long response time of industrial load but large regulation range, rapid response of energy storage and strong two-way regulation ability, and construct an operating characteristic assessment index. Through multi-objective correlation analysis, it is clear that photovoltaic and wind power are suitable for undertaking the basic output task, energy storage is suitable for performing rapid frequency modulation and peak-valley filling tasks, and industrial load is suitable for demand response and standby tasks, forming a resource-task matching degree matrix. According to the matching degree matrix, the resources are divided into three functional areas: basic power supply area, rapid regulation area, and emergency response area, and an inter-regional mutual assistance strategy is formulated. Based on the functional area division and load forecasting, the optimal resource allocation plan for each time period is calculated. For example, during the morning load ramp-up period, it mainly relies on energy storage discharge and wind power peak shaving, and during the peak photovoltaic output period at noon, industrial loads are guided to increase consumption. Finally, through smoothing processing in the time dimension and conflict resolution in the space dimension, a collaborative scheduling strategy for the virtual power plant group covering the entire region for 24 hours a day is formed to achieve the economic and efficient operation of the power system.

[0087] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0088] (1) Conduct historical trend analysis and fault feature extraction on the equipment health status assessment data to generate an equipment health degradation curve;

[0089] (2) Conduct correlation analysis between the equipment health degradation curve and the equipment scheduling load in the collaborative scheduling strategy of the virtual power plant group to obtain an equipment operation pressure index;

[0090] (3) Based on the equipment operation pressure index, predict the remaining life and risk rating of the equipment, and construct an equipment status warning table;

[0091] (4) Classify and sort the equipment in the equipment status warning table according to importance and operation risk to generate a maintenance priority matrix;

[0092] (5) According to the maintenance priority matrix and the collaborative scheduling strategy of the virtual power plant group, divide the maintenance time window to obtain a maintenance time sequence arrangement table;

[0093] (6) Configure maintenance resources and plan maintenance processes for the maintenance schedule to obtain the critical equipment maintenance plan.

[0094] Specifically, through in-depth analysis of the equipment health status evaluation data, extract the key features during the operation of the equipment. Conduct time series analysis on historical data, including the change trend of electrical parameters of the equipment, temperature change law, vibration characteristics, etc. Each characteristic parameter has its unique change pattern. For example, the winding temperature of a transformer shows periodic changes during normal operation, while it shows obvious mutation characteristics when abnormal. By extracting and analyzing these features, draw a health decline curve reflecting the change of equipment health status over time. After obtaining the equipment health decline curve, it is necessary to conduct correlation analysis with the load data in the coordinated dispatching strategy of the virtual power plant group. The dispatching load data contains the operation condition information of the equipment, such as load rate, start-stop frequency, load change rate, etc. By analyzing the relationship between the equipment health status and the operation condition, calculate the equipment operation pressure index. This index comprehensively reflects the matching degree between the equipment health status and the operation load.

[0095] Based on the equipment operation pressure index, predict the remaining life of the equipment. Consider the historical operation data, current health status and future expected operation conditions of the equipment during the prediction process. At the same time, conduct risk rating on the equipment according to the prediction results. The rating criteria include multiple dimensions such as equipment importance, failure probability, and failure impact range. Through these analyses, construct an equipment status warning table containing various risk indicators. The equipment status warning table contains a large amount of risk assessment information of equipment. Analyze and sort this information, focusing on the two dimensions of equipment importance in the power grid and operation risk. The equipment importance is determined by factors such as its location in the power grid, the functions it undertakes, and the influence range. The operation risk is based on the previous health status evaluation and remaining life prediction results. Through comprehensive analysis of these factors, generate a maintenance priority matrix.

[0096] According to the sorting result of the maintenance priority matrix, combined with the coordinated dispatching strategy of the virtual power plant group, reasonably arrange the equipment maintenance time. The dispatching strategy provides load prediction and resource allocation information for different time periods. By analyzing this information, select a suitable maintenance time window. The selection of the time window needs to balance the urgency of maintenance and the impact on system operation, and divide to obtain the maintenance schedule. Further refine the maintenance schedule, configure the required maintenance resources, including personnel, materials, equipment, etc. At the same time, plan specific maintenance processes, including pre-maintenance preparation, in-maintenance process control, post-maintenance acceptance, etc. Through these detailed plans, finally form a complete critical equipment maintenance plan.

[0097] Taking the overhaul plan of the main transformer of a certain substation as an example: Through the analysis of the operation data of this transformer in the past year, it is found that its temperature data shows a gradually rising trend, the vibration amplitude of the iron core increases, and the content of dissolved gases in the oil is abnormal. The health degradation curve drawn based on the data shows that the health state of the equipment is deteriorating at an accelerating rate. At the same time, this transformer undertakes the important task of supplying industrial loads, with an average load rate reaching 85% and frequent load adjustments. The operation pressure index obtained through correlation analysis is relatively high, and the predicted remaining life is lower than expected, so it is classified as a high-risk level. Considering the important position of the transformer in the power grid and its current health state, it is listed at the top of the overhaul priority list. Combining the load forecast, it is found that the industrial load drops significantly during the Spring Festival. Selecting this time window for overhaul, a detailed overhaul plan including winding inspection, oil quality treatment, iron core fastening, etc. is formulated.

[0098] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0099] (1) Decompose the collaborative scheduling strategy of the virtual power plant group according to four time scales: annual, monthly, day-ahead, and real-time, to obtain a hierarchical scheduling strategy matrix;

[0100] (2) Map the overhaul plan of key equipment in the time dimension to generate a time series table of equipment available capacity;

[0101] (3) Conduct a matching analysis between the hierarchical scheduling strategy matrix and the time series table of equipment available capacity to construct a set of system operation constraint boundaries;

[0102] (4) Configure and divide the operating parameters by time period according to the set of system operation constraint boundaries to obtain a parameter adjustment range table;

[0103] (5) Conduct a time series coupling analysis and mutual feedback impact assessment on the parameters in the parameter adjustment range table to generate a parameter coordination control sequence;

[0104] (6) Integrate and optimize the parameter coordination control sequence on multiple time scales to obtain a power system balance control scheme.

[0105] Specifically, such as Figure 2As shown in the figure, it is a schematic flowchart of multi-time-scale optimal configuration of system operation parameters in the embodiments of the present application. The decomposition of the collaborative scheduling strategy for the virtual power plant group starts from the annual scale. The annual scheduling mainly focuses on the equipment maintenance cycle, seasonal load characteristics, and renewable energy output patterns. The monthly scheduling focuses on the load change trend within the month and the equipment maintenance arrangement. The day-ahead scheduling focuses on the load prediction and resource allocation within 24 hours. The real-time scheduling adjusts for minute-level power balance. Through this hierarchical decomposition, a hierarchical scheduling strategy matrix containing scheduling requirements at different time scales is obtained. When mapping the key equipment maintenance plan in the time dimension, the impact of the maintenance work on the available capacity of the equipment needs to be considered. The maintenance process of the equipment usually includes two methods: full-stop maintenance and on-load maintenance, and the corresponding capacity impacts are also different. By analyzing the maintenance procedures and durations, the maintenance plan is converted into time series data of the available capacity of the equipment, and a time series table of the available capacity of the equipment is generated.

[0106] When performing a matching analysis between the hierarchical scheduling strategy matrix and the time series table of the available capacity of the equipment, it is necessary to identify the relationship between the capacity constraints and scheduling requirements at each time scale. At the annual scale, the matching between the maintenance plan and the seasonal load is mainly considered. At the monthly and day-ahead scales, the focus is on the impact of maintenance activities on scheduling resources. At the real-time scale, the temporary constraints during the equipment maintenance process need to be considered. Through these analyses, a complete set of system operation constraint boundaries is constructed. Based on the system operation constraint boundary set, the system operation parameters are configured in time periods and divided into intervals. The operation parameters include key indicators such as voltage, frequency, and power, and each parameter has its reasonable operation interval. By analyzing the load characteristics and network states at different time periods, the adjustment range of each parameter is determined, and a parameter adjustment range table is formed.

[0107] Perform a time series coupling analysis on the parameters in the parameter adjustment range table, focusing on studying the mutual influence relationship between the parameters. For example, power adjustment will affect the voltage level, and frequency change will affect the equipment operation state. By analyzing these coupling relationships, the mutual feedback impact of parameter adjustment is evaluated, and a parameter coordination control sequence is generated. Integrate and optimize the parameter coordination control sequence at different time scales. The optimization at the annual level mainly considers the coordination between equipment maintenance and seasonal load. The monthly level focuses on optimizing the coordination between the maintenance plan and monthly scheduling. The day-ahead level focuses on the coping strategies for intra-day load fluctuations. The real-time level mainly solves the instantaneous power balance problem. Through the collaborative optimization of multiple time scales, a complete power system balance control scheme is finally formed.

[0108] Taking a distribution network area as an example, this area contains multiple distributed power sources and controllable loads. Annual scheduling analysis shows that the load is higher in summer and lower in winter, while distributed photovoltaic power generation shows the opposite seasonal characteristics. Monthly analysis finds that the industrial load is larger in the middle of each month and relatively smaller at the beginning and end of the month. Day-ahead analysis indicates that there are obvious load peaks in the morning and evening on weekdays. Real-time data shows that the fluctuations of distributed photovoltaic power generation have a greater impact on system stability. Based on these characteristics, a hierarchical scheduling strategy is formulated: at the annual level, the maintenance of large equipment is arranged during the low-load period in winter; at the monthly level, daily maintenance is arranged at the beginning or end of the month; day-ahead scheduling makes full use of the energy storage system to cope with the morning and evening peaks; real-time scheduling suppresses the photovoltaic fluctuations through a fast-response energy storage device. At the same time, by analyzing the constraint conditions at each time scale, the adjustment ranges of operating parameters such as voltage qualification rate and power factor are determined. The finally formed balance control scheme not only meets the operating requirements at different time scales but also ensures the safe and stable operation of the system.

[0109] In a specific embodiment, the process of performing the step of analyzing the matching between the hierarchical scheduling strategy matrix and the equipment available capacity time series table may specifically include the following steps:

[0110] (1) Perform time series analysis on the scheduling data of each time scale in the hierarchical scheduling strategy matrix to generate a scheduling strategy time series feature table;

[0111] (2) Perform capacity balance analysis on the equipment available capacity time series table according to time nodes to obtain the equipment operation availability curve;

[0112] (3) Calculate the resource matching degree according to the scheduling strategy time series feature table and the equipment operation availability curve, and construct a resource allocation constraint matrix;

[0113] (4) Perform voltage constraint, power constraint, and capacity constraint analysis on the resource allocation constraint matrix to form a set of system operation boundary conditions;

[0114] (5) Perform regional coordination evaluation on the set of system operation boundary conditions according to spatial distribution to obtain a regional mutual assistance constraint table;

[0115] (6) Integrate the boundary conditions and perform constraint mapping on the regional mutual assistance constraint table to generate a set of system operation constraint boundaries.

[0116] Specifically, the hierarchical scheduling strategy matrix contains scheduling data for four time scales: annual, monthly, day-ahead, and real-time. In the time series analysis process, the scheduling data for each time scale is first unified. The annual data is decomposed monthly, the monthly data is decomposed daily, the day-ahead data is decomposed hourly, and the real-time data is decomposed by minute, forming a scheduling strategy time series with a unified time granularity. Time series decomposition technology is used for time series analysis, decomposing the scheduling data for each time scale into three components: trend term, periodic term, and random term, and extracting key time series features, such as time series characteristic parameters like peak value, valley value, ramp rate, and duration. Finally, a scheduling strategy time series feature table is formed. This feature table has time as rows and feature indicators as columns, clearly showing the scheduling characteristics at each time node. When performing capacity balance analysis on the device available capacity time series table according to time nodes, it is necessary to perform aggregation and balance calculations in the time dimension for the available capacities of various types of power generation equipment, energy storage equipment, and adjustable loads. First, the equipment is grouped by type, and the total available capacity of each type of equipment at different time nodes is calculated and summarized. Then, according to the power balance principle, the supply-demand balance state of the system at each time node is calculated. The capacity balance analysis uses the sliding time window method, with a specific time window (such as 30 minutes, 1 hour) as the basic unit, calculating the capacity balance index within the window and forming a continuous curve of device operation availability. This curve intuitively shows the change in the capacity available state of the system on the time axis, facilitating the identification of capacity shortage periods and surplus periods.

[0117] When calculating the resource matching degree based on the scheduling strategy time series feature table and the device operation availability curve, a spatio-temporal matching algorithm is used to perform corresponding analysis on the scheduling requirements and resource capabilities. For each time node, the demand characteristics of the scheduling strategy (such as frequency regulation demand, reserve demand, peak shaving demand, etc.) are matched with the available capacity of the device at the corresponding moment to evaluate the matching degree, and the adaptation degree of various resources to various scheduling requirements is calculated. The matching degree calculation considers multi-dimensional factors such as resource response time, regulation rate, and continuous ability, and uses a weighted scoring method to comprehensively evaluate the adaptation degree of resources to specific scheduling requirements, constructing a resource allocation constraint matrix. This matrix has resource types as rows and time nodes as columns, and the matrix element values represent the allocation constraint conditions of specific resources at specific moments, such as the maximum schedulable capacity, the minimum capacity that must be reserved, etc.

[0118] When analyzing voltage constraints, power constraints, and capacity constraints on the resource allocation constraint matrix, it is necessary to comprehensively consider the physical characteristics and safe operation limits of the power system. Voltage constraint analysis is based on the results of power flow calculations to extract the allowable voltage change ranges of each node and form a voltage constraint set; power constraint analysis is based on the rated parameters of equipment and system stability requirements to extract the power transmission limits of each transmission line and equipment and form a power constraint set; capacity constraint analysis is based on equipment reliability and scheduling flexibility requirements to extract the capacity scheduling limits of various resources and form a capacity constraint set. By integrating these three types of constraint sets, a complete set of system operation boundary conditions is formed, which describes the various constraint conditions for the safe and stable operation of the system in a standardized format. When evaluating the regional coordination of the system operation boundary condition set according to spatial distribution, based on the geographical division of the power system, the power transmission capacity, resource complementarity, and emergency mutual assistance capabilities among regions are comprehensively evaluated. The regional coordination evaluation uses the regional coupling degree analysis method to calculate the power flow capacity and resource complementarity degree among regions, identify key interconnection channels and regional cooperation bottlenecks, and form a regional mutual assistance constraint table. This constraint table clearly shows the mutual assistance capacity boundaries among regions, including key indicators such as the maximum transmission capacity, minimum reserved capacity, and emergency mutual assistance response time among regions.

[0119] When integrating boundary conditions and constraint mapping for the regional mutual assistance constraint table, the mutual assistance constraints at the regional level are integrated with the operation constraints at the equipment level to form a multi-level and full-dimensional constraint system. The boundary condition integration uses a constraint fusion algorithm to uniformly express constraint conditions at different levels and dimensions, eliminate redundant constraints, strengthen key constraints, and form a concise and comprehensive constraint set. Constraint mapping is to map abstract constraint conditions to specific control parameters and scheduling variables, clarify the control objectives and limit ranges corresponding to each constraint condition, and finally generate a system operation constraint boundary set, which provides a clear operation boundary for the subsequent power system balance control scheme.

[0120] Taking a certain regional power grid as an example, this power grid includes various power generation resources such as thermal power, hydropower, wind power, and photovoltaic power, as well as flexible regulation resources such as pumped storage and battery energy storage. First, perform a time series analysis on the hierarchical scheduling strategy matrix, decompose the annual scheduling strategy plan for the seasonal regulation of hydropower to the monthly granularity, decompose the unit maintenance arrangement in the monthly scheduling strategy to the daily granularity, and decompose the unit start-stop in the day-ahead scheduling strategy to the hourly granularity to form a time series feature table containing the scheduling characteristics of each time node. This table shows that during the peak heating load period in winter, the scheduling strategy requires thermal power units to operate at full capacity, while during the spring flood season, it requires hydropower to generate electricity first. Then, generate a time series table of equipment available capacity according to the key equipment maintenance plan. Through sliding time window analysis, it is found that in the third week of a certain month, due to the simultaneous maintenance of two large thermal power units, the system reserve capacity drops below the warning line, forming a "valley" in the equipment operation availability curve. Based on the scheduling strategy time series feature table and the equipment operation availability curve, calculate the matching degree of various resources. For example, gas turbine units and battery energy storage with fast response times have a high matching degree in real-time frequency modulation, while pumped storage with a large adjustment range has a high matching degree in intraday peak shaving, and construct a resource allocation constraint matrix containing spatio-temporal dimension constraint information. Perform multi-dimensional constraint analysis on this matrix to determine the power flow limit of each transmission channel, the voltage qualification range of each bus, and the scheduling capacity limit of various resources to form a set of system operation boundary conditions. According to the regional division of the system, evaluate the coordination between regions. For example, Region A is rich in wind power resources but has limited consumption capacity, forming a mutual assistance relationship with Region B where the electricity load is concentrated. However, due to the limited capacity of the tie line, the mutual assistance capacity has an upper limit, and generate a detailed regional mutual assistance constraint table. Finally, integrate the constraint conditions at the equipment level and the regional level, clarify the control objectives corresponding to each constraint, such as the tie line power flow not exceeding 80% of the rated capacity, the voltage deviation of each bus not exceeding ±5% of the rated value, and the minimum spinning reserve capacity of each region not less than 50% of the largest unit capacity, to form a set of system operation constraint boundaries, providing clear boundary conditions for the formulation of the power system balance control plan.

[0121] The above describes the power system planning operation optimization method in the embodiments of the present application. Next, the power system planning operation optimization system in the embodiments of the present application will be described. Please refer to Figure 3 , an embodiment of the power system planning operation optimization system in the embodiments of the present application includes:

[0122] Through the collaborative cooperation of the above-mentioned various components, by collecting and extracting features in multiple dimensions for electrical parameters, equipment status parameters, and environmental parameters, a comprehensive equipment health status evaluation system has been established, improving the accuracy and timeliness of equipment status monitoring; based on the method of resource grouping and priority division using equipment health status evaluation data and regional extreme weather warning information, the dynamic optimal allocation of virtual power plant resources has been achieved, enhancing the system's adaptability to extreme weather; by globally decomposing tasks and allocating targets for the capacity distribution and operation characteristics of the virtual power plant resource pool, a method for generating a scheduling strategy based on multi-level collaboration has been proposed, improving the coordinated control effect of the system; according to the equipment health status evaluation data and the collaborative scheduling strategy of the virtual power plant group, state prediction and maintenance timing arrangement of the equipment have been carried out, constructing a preventive maintenance decision-making mechanism, reducing the equipment failure risk; based on the collaborative scheduling strategy of the virtual power plant group and the overhaul plan of key equipment, multi-time scale optimal allocation of system operation parameters has been carried out, designing a dynamic balance control scheme, achieving the improvement of system operation efficiency. In the data processing aspect of the present invention, a deep learning feature fusion model is adopted, which can effectively extract the key features of the equipment operation state, dynamically weight different features through the attention mechanism, improving the accuracy of state evaluation; in the resource scheduling aspect, a graph neural network algorithm is applied for dynamic grouping, which fully considers the topological relationship and operation characteristics among resources, enhancing the collaborative control effect of the virtual power plant; in the optimization decision-making aspect, a multi-agent reinforcement learning method is introduced to achieve the adaptive optimal control of the system, enhancing the practicality and popularizability of the solution.

[0123] Referring to Figure 4 , in the embodiment of the present invention, a computer device is further provided. This computer device can be a server, and its internal structure can be as Figure 4 shown. This computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor of this computer design is used to provide computing and control capabilities. The memory of this computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of this computer device is used to store the corresponding data in this embodiment. The network interface of this computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0124] Those skilled in the art can understand that Figure 4 the structure shown in

[0125] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0126] Those of ordinary skill in the art can understand that all or part of the processes in the above-described method embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above-described method embodiments. Among them, any reference to a memory, storage, database, or other medium provided by the present invention and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0127] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0128] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0129] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of various embodiments of the present application.

Claims

1. A method for optimizing the planning and operation of a power system, characterized in that, Including multi-dimensional acquisition and feature extraction of power system operation data based on electrical parameters, equipment status parameters, and environmental parameters to obtain equipment health status evaluation data. The electrical parameters include voltage, current, active power, and reactive power. The equipment status parameters include temperature, vibration, noise, and partial discharge data. The environmental parameters include air temperature, humidity, wind speed, and rainfall data; Based on the equipment health status evaluation data and regional extreme weather warning information, perform resource grouping and priority division on distributed energy, controllable loads, and energy storage systems to form a virtual power plant resource pool; According to the capacity distribution and operation characteristics of the virtual power plant resource pool, perform global task decomposition and target allocation to generate a collaborative scheduling strategy for the virtual power plant group; Based on the equipment health status evaluation data and the collaborative scheduling strategy of the virtual power plant group, perform state prediction and maintenance time sequence arrangement for the equipment to obtain the maintenance plan for key equipment; Based on the collaborative scheduling strategy of the virtual power plant group and the maintenance plan for key equipment, perform multi-time scale optimal configuration of the system operation parameters to obtain a power system balance control plan, including decomposing the collaborative scheduling strategy of the virtual power plant group according to four time scales: annual, monthly, day-ahead, and real-time, to obtain a hierarchical scheduling strategy matrix; performing time dimension mapping on the maintenance plan for key equipment to generate a time series table of equipment available capacity; Perform matching analysis on the hierarchical scheduling strategy matrix and the time series table of equipment available capacity to construct a system operation constraint boundary set, including performing time series analysis on the scheduling data of each time scale in the hierarchical scheduling strategy matrix to generate a scheduling strategy time series feature table; performing capacity balance analysis on the time series table of equipment available capacity according to time nodes to obtain an equipment operation availability curve; calculating the resource matching degree according to the scheduling strategy time series feature table and the equipment operation availability curve to construct a resource allocation constraint matrix; performing voltage constraint, power constraint, and capacity constraint analysis on the resource allocation constraint matrix to form a system operation boundary condition set, performing boundary condition integration and constraint mapping on the regional mutual assistance constraint table to generate the system operation constraint boundary set; performing regional coordination evaluation on the system operation boundary condition set according to spatial distribution to obtain a regional mutual assistance constraint table; According to the system operation constraint boundary set, perform sub-period configuration and interval division on the operation parameters to obtain a parameter adjustment range table; perform time series coupling analysis and mutual feedback impact evaluation on the parameters in the parameter adjustment range table to generate a parameter coordination control sequence; perform multi-time scale integration and optimization processing on the parameter coordination control sequence to obtain a power system balance control plan.

2. The power system planning and operation optimization method according to claim 1, characterized in that Multi-dimensionally collect and extract features from the operation data of the power system according to electrical parameters, equipment status parameters, and environmental parameters to obtain equipment health status evaluation data, including: collecting voltage, current, active power, and reactive power data in electrical parameters, performing standardization processing on the collected data to obtain electrical feature vectors; collecting equipment temperature, vibration, noise, and partial discharge data in equipment status parameters, performing time-frequency domain analysis on the collected data through wavelet transform to obtain equipment status feature vectors; collecting air temperature, humidity, wind speed, and rainfall data in environmental parameters, performing normalization processing on the collected data to obtain environmental feature vectors; performing feature fusion on the electrical feature vectors, equipment status feature vectors, and environmental feature vectors to construct an equipment operation feature matrix; performing time series correlation analysis on the dynamic data in the equipment operation feature matrix to obtain the equipment dynamic health index; performing life cycle assessment on the static data in the equipment operation feature matrix to obtain the equipment static health index; performing weighted fusion on the equipment dynamic health index and the equipment static health index to obtain the equipment health status evaluation data.

3. The power system planning and operation optimization method according to claim 1, characterized in that Based on the equipment health status evaluation data and regional extreme weather warning information, group resources and divide priorities for distributed energy, controllable loads, and energy storage systems to form a virtual power plant resource pool, including: analyzing the equipment health status evaluation data, extracting the health risk coefficient of each equipment, and generating an equipment vulnerability evaluation matrix; identifying the types of weather disasters and dividing the intensity levels according to the regional extreme weather warning information to obtain a set of weather impact factors, and performing correlation analysis between the equipment vulnerability evaluation matrix and the set of weather impact factors to construct an equipment-weather sensitivity table; performing regional clustering on distributed energy, controllable loads, and energy storage systems based on the equipment-weather sensitivity table to obtain a resource grouping sequence, and performing available capacity evaluation and response ability analysis on various resources in the resource grouping sequence to generate resource regulation priority indicators; performing geographical location correlation analysis on each group of resources in the resource grouping sequence to generate a resource spatial distribution matrix; performing cross-mapping between the resource spatial distribution matrix and the resource regulation priority indicators to obtain an initial resource combination table; performing energy complementarity analysis on the resources in the initial resource combination table to construct a resource complementarity matrix, and optimizing and reorganizing the initial resource combination table according to the resource complementarity matrix to obtain a resource coordination index; performing weighted fusion on the resource coordination index and the resource regulation priority indicators to form a virtual power plant resource pool.

4. The power system planning and operation optimization method according to claim 1, wherein Perform global task decomposition and target allocation according to the capacity distribution and operation characteristics of the virtual power plant resource pool, and generate a collaborative scheduling strategy for the virtual power plant group, including: hierarchically statistically analyzing the capacity distribution data in the virtual power plant resource pool according to energy types and capacity sizes, extracting spatio-temporal distribution characteristics, and generating a resource capacity basic matrix; performing volatility analysis and stability evaluation on the data in the resource capacity basic matrix, and comparing with historical operation data to obtain a capacity availability evaluation table; performing multi-dimensional parameter extraction and feature mapping according to the output characteristics, response time, and regulation ability of various resources in the virtual power plant resource pool, and constructing an operation characteristic evaluation index; performing multi-objective correlation analysis on the capacity availability evaluation table and the operation characteristic evaluation index to obtain a resource-task fitness matrix; performing hierarchical decomposition and boundary identification on the resource-task fitness matrix to generate a task partition table and a cross-region mutual assistance plan; performing resource allocation and load distribution based on the task partition table and the cross-region mutual assistance plan to obtain an initial scheduling instruction set; and performing collaborative optimization and conflict resolution on the initial scheduling instruction set according to time series and spatial regions to generate a collaborative scheduling strategy for the virtual power plant group.

5. The power system planning and operation optimization method according to claim 1, wherein Based on the equipment health status evaluation data and the collaborative scheduling strategy of the virtual power plant group, perform equipment status prediction and maintenance timing arrangement to obtain a key equipment maintenance plan, including: performing historical trend analysis and fault feature extraction on the equipment health status evaluation data to generate an equipment health degradation curve; performing correlation analysis between the equipment health degradation curve and the equipment scheduling load in the collaborative scheduling strategy of the virtual power plant group to obtain an equipment operation pressure index; predicting the remaining life and risk rating of the equipment based on the equipment operation pressure index, and constructing an equipment status warning table; classifying and sorting the equipment in the equipment status warning table according to importance and operation risk to generate a maintenance priority matrix; dividing the maintenance time window according to the maintenance priority matrix and the collaborative scheduling strategy of the virtual power plant group to obtain a maintenance timing arrangement table; and performing maintenance resource configuration and maintenance process planning on the maintenance timing arrangement table to obtain a key equipment maintenance plan.

6. A power system planning and operation optimization system for implementing the power system planning and operation optimization method according to any one of claims 1-5, characterized in that, Including: An extraction module for multi-dimensionally collecting and extracting features from the power system operation data according to electrical parameters, equipment status parameters, and environmental parameters to obtain equipment health status evaluation data; A division module for grouping resources and dividing priorities for distributed energy, controllable loads, and energy storage systems based on the equipment health status evaluation data and regional extreme weather warning information to form a virtual power plant resource pool; An allocation module for performing global task decomposition and target allocation according to the capacity distribution and operation characteristics of the virtual power plant resource pool to generate a collaborative scheduling strategy for the virtual power plant group; A prediction module for predicting the equipment status and arranging the maintenance timing based on the equipment health status evaluation data and the collaborative scheduling strategy of the virtual power plant group to obtain a key equipment maintenance plan; A configuration module for multi-time scale optimal configuration of the system operation parameters based on the collaborative scheduling strategy of the virtual power plant group and the key equipment maintenance plan to obtain a power system balance control plan.

7. A computer device, characterized in that, It includes a memory and a processor. The memory stores a computer program that can run on the processor. It is characterized in that when the processor executes the computer program, it implements the power system planning operation optimization method described in any one of claims 1 to 5.

8. A computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, the processor is caused to execute the power system planning operation optimization method described in any one of claims 1 to 5.

Images