Power grid collaborative optimization operation method for distributed power supply power injection

Abstract

The invention relates to the technical field of power grid operation and maintenance, provides a power grid collaborative optimization operation method for distributed power supply power injection, and aims to realize efficient and stable operation of a power grid by combining a genetic algorithm, a power grid simulation model and a control model. And obtaining a power injection scheme which meets the power grid constraint and has an optimal target. Meanwhile, the power grid simulation model is used for conducting simulation evaluation on the scheme, the practicability is improved, the control model rapidly adjusts the power injection scheme according to the real-time power grid state and task requirements and is applied to an actual power grid, the dispatching flexibility of the power grid is improved, efficient utilization of the distributed power supply is promoted, and the operation cost of the power grid is reduced; and the adaptability of the power grid is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid operation and maintenance, and in particular to a method for coordinated optimization operation of a power grid with distributed power generation injection. Background Art

[0002] Grid collaborative optimization of distributed power generation power injection refers to the effective collaboration between distributed power sources (such as solar photovoltaic, wind power, etc.) and the main power grid through control strategies and technical means. The optimization process aims to improve the operating efficiency, stability and safety of the power grid. Specifically, grid collaborative optimization of distributed power generation power injection includes the regulation of active power and reactive power of distributed power sources to achieve optimal control of performance indicators such as grid voltage and power factor. Through hierarchical control strategies, multi-agent systems and other technologies, the power injection of distributed power access points can be controlled so that they can track the reference instructions of the main power grid dispatching center and achieve active power sharing and voltage balancing between distributed power sources.

[0003] In the collaborative optimization process of distributed power generation injection into the power grid, there are the following technical pain points: the output power of distributed power generation is affected by factors such as weather and operating conditions and shows strong volatility and intermittency, which increases the difficulty of power grid frequency and voltage control, making traditional deterministic planning difficult to adapt to the construction of new power systems. In order to solve this technical pain point, the present invention provides a method for collaborative optimization operation of a power grid with distributed power generation injection. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a method for coordinated optimization operation of a power grid with distributed power generation injection, so as to solve the problem that the output power of distributed power generation is affected by factors such as weather and operating conditions, and exhibits strong volatility and intermittency, which increases the difficulty of controlling the frequency and voltage of the power grid and makes it difficult for traditional deterministic planning to adapt to the construction of new power systems.

[0005] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0006] The method for coordinated optimization operation of a power grid with distributed power generation power injection of the present invention comprises:

[0007] Step S101, obtaining real-time power data of distributed power sources, power grid status data, power grid topology data, distributed power source control strategy data, power grid dispatch center command data and historical operation data;

[0008] Step S102, obtaining natural environment data of the location where the power grid is located, performing cluster analysis on the natural environment data of the location where the power grid is located, obtaining a clustering result of the natural environment of the location where the power grid is located, establishing an association relationship between the clustering result of the natural environment of the location where the power grid is located and time information, and obtaining a knowledge base of the natural environment of the power grid;

[0009] Step S103, receiving a grid task for distributed power generation power injection, the grid task for distributed power generation power injection including basic task information, distributed power generation information and grid constraint data, matching the grid task for distributed power generation power injection in a grid natural environment knowledge base to obtain natural environment data corresponding to the grid task for distributed power generation power injection;

[0010] Step S104, using the real-time power data of distributed power sources, the power grid state data, the power grid topology data, the distributed power source control strategy data, the power grid dispatch center command data and the historical operation data to train the random forest model, to obtain a power grid simulation model of distributed power source power injection, and to substitute the basic task information into the power grid simulation model of distributed power source power injection, to obtain a power grid simulation result of distributed power source power injection;

[0011] Step S105, compare the grid simulation result of distributed power generation power injection with the grid constraint data to obtain the comparison result of the grid simulation result of distributed power generation power injection; if the grid simulation result of distributed power generation power injection is outside the range of the grid constraint data, then optimize the grid simulation model of distributed power generation power injection using a genetic algorithm based on the real-time power data of the distributed power generation, the grid status data and the grid topology data to obtain an optimized grid simulation model of distributed power generation power injection; use the optimized grid simulation model of distributed power generation power injection as the grid control model of distributed power generation power injection; substitute the grid task of distributed power generation power injection into the grid control model of distributed power generation power injection to obtain real-time distributed power generation power injection grid control parameters.

[0012] Furthermore, the method for coordinated optimization operation of a power grid with distributed power generation power injection according to the present invention, step S101 comprises:

[0013] Through the interface with the monitoring system or data acquisition and monitoring control system of distributed power sources, the current output power data of each distributed power source is obtained in real time, and the real-time power data of the distributed power source is obtained. The distributed power source includes a solar power station, a wind power station and an energy storage system;

[0014] Acquire the real-time status data of the power grid from the monitoring system of the power grid. The real-time status data of the power grid includes the voltage, current, frequency, power factor of each node, the load condition of the power grid, and the line condition. The line condition includes the line condition overload and line fault, and obtain the power grid status data;

[0015] The topological structure information of the power grid includes the connection relationship and parameters of the nodes, lines, transformers, and switchgear of the power grid;

[0016] The control strategy information of the distributed power source includes power control strategy, voltage control strategy and frequency control strategy. The power control strategy includes constant power control and maximum power point tracking control.

[0017] Obtain dispatching instructions for distributed power sources from the power grid dispatching center, including power dispatching instructions, start and stop instructions, and reactive power compensation instructions, which are used to coordinate the power exchange between distributed power sources and the power grid and obtain the command data of the power grid dispatching center;

[0018] The historical operation data is obtained from the historical database of the power grid and distributed power sources. The historical operation data includes past power data, power grid status data, fault records, and maintenance records.

[0019] Furthermore, the method for coordinated optimization operation of a power grid with distributed power generation power injection according to the present invention, step S102 comprises:

[0020] Obtaining natural environment data of the location where the power grid is located. The natural environment data of the location where the power grid is located includes meteorological data, geographical data and environmental data. Meteorological data includes temperature, humidity, wind speed, wind direction, precipitation and air pressure. Geographic data includes topography, altitude and soil type. Environmental data includes air quality and solar radiation intensity.

[0021] Preprocess the natural environment data, extract features from the preprocessed natural environment data, and obtain meteorological features to reflect the impact of the natural environment on the operation of the power grid. The meteorological features include daily average temperature features, maximum temperature features, minimum temperature features, wind speed mean features, and precipitation features;

[0022] Use a clustering algorithm to perform cluster analysis on the extracted features to obtain natural environment clustering results, and establish an association relationship between the natural environment clustering results and the corresponding time information;

[0023] Integrate natural environment data, natural environment clustering results and their association with time information into a structured knowledge base.

[0024] Furthermore, the method for coordinated optimization operation of a power grid with distributed power generation power injection according to the present invention, step S103 comprises:

[0025] Receive the grid task injected by the distributed power generation, parse the received grid task data, and extract the basic task information, distributed power generation information and grid constraint data;

[0026] The basic information of the task includes the task ID, task name, planned execution time and task priority;

[0027] Distributed power information includes the type, rated power and access point location of distributed power;

[0028] Grid constraint data includes the voltage level, frequency stability requirements, and power factor limits of the grid;

[0029] Extracting task execution time information from the task basic information, the task execution time information includes the date and time period of the planned execution time, using the task execution time information to search for natural environment data clustering results corresponding to the task execution time in the power grid natural environment knowledge base, the matching process involves searching for natural environment data corresponding to a specific date, time period or season;

[0030] Extracting natural environment data matching the task execution time from the power grid natural environment knowledge base, the natural environment data including temperature, wind speed, precipitation, light information and wind speed information during the period;

[0031] The acquired natural environment data is integrated with the grid mission data injected by distributed generation power, and the integrated data is used for subsequent grid simulation, optimization and control processes.

[0032] Furthermore, the method for coordinated optimization operation of a power grid with distributed power generation power injection according to the present invention, step S104 comprises:

[0033] Summarize the real-time power data of distributed power sources, power grid status data, power grid topology data, distributed power source control strategy data, power grid dispatch center command data and historical operation data;

[0034] Preprocess the data and extract features from the preprocessed data to obtain the real-time power of the distributed power source, the voltage of each node in the power grid, the current of each node in the power grid, the power flow of the line, and the control parameters of the distributed power source;

[0035] The data set is divided into a training set and a test set. The training set is used to train the random forest model, and the test set is used to evaluate the performance of the model.

[0036] Use the training set data to train the random forest model, receive and set the parameters of the random forest, the parameters of the random forest include the number of trees, the maximum depth, and the minimum number of sample splits, run the random forest algorithm, obtain the power grid simulation model of distributed power generation injection, substitute the basic information of the power grid task of distributed power generation injection into the power grid simulation model of distributed power generation injection, the basic information of the power grid task includes the task time, the type of distributed power generation and the power data, run the power grid simulation model of distributed power generation injection, and obtain the power grid simulation results of distributed power generation injection, which include the voltage changes of each node of the power grid, the power flow of the line and the power output data of the distributed power generation.

[0037] Furthermore, the method for coordinated optimization operation of a power grid with distributed power generation power injection according to the present invention, step S105 comprises:

[0038] Determine the optimization target, based on the actual needs and constraints of the power grid, to determine the optimization target, which includes minimizing the loss of the power grid, stabilizing the voltage, and maximizing the utilization of distributed power sources;

[0039] Receive and set basic parameters of the genetic algorithm, including population size, genetic generation, crossover probability, and mutation probability;

[0040] The power injection scheme of the distributed generation is represented as an individual in the genetic algorithm by using real number coding, and the individual includes the power setting value or control parameter of the distributed generation, and the individual is taken as the object of optimization of the genetic algorithm;

[0041] Initialize the population. According to the set population size, randomly generate a group of initial individuals to form the initial population. Each individual represents a distributed power injection scheme.

[0042] Using the real-time power data of distributed generation, grid status data and grid topology data, the fitness of each individual is evaluated through the grid simulation model. The fitness function should be designed according to the optimization goal.

[0043] According to the fitness evaluation results, individuals with higher fitness are selected as parents for crossover and mutation operations;

[0044] Perform crossover operation on the selected parent individuals to generate new offspring individuals;

[0045] The crossover operation is performed by single-point crossover, double-point crossover or uniform crossover;

[0046] Perform mutation operations on offspring individuals by changing the gene values ​​in the individuals or introducing new genes;

[0047] Update the population, replace some or all of the parent individuals with the generated offspring individuals, update the population, and determine whether to continue iterating based on the set genetic generation or convergence conditions;

[0048] If the set genetic generation number is reached or the fitness meets the convergence condition, the iteration is stopped. After iterative optimization of the genetic algorithm, the individual with the highest fitness is obtained, that is, the optimized distributed power generation power injection scheme, which is used as the parameter of the optimized distributed power generation power injection power grid simulation model.

[0049] Furthermore, the method for coordinated optimization operation of a power grid with distributed power generation power injection according to the present invention, step S105 comprises:

[0050] The power grid simulation model of distributed power generation power injection is optimized by genetic algorithm, and the optimized model is obtained, and the parameters of the optimized power grid simulation model of distributed power generation power injection are set as the initial parameters of the power grid control model of distributed power generation power injection;

[0051] Receive the grid task of distributed generation power injection from the grid dispatch center. The grid task should include basic information of the task, such as task time, distributed generation type and expected power injection amount;

[0052] Preprocessing the received power grid task data, substituting the preprocessed power grid task data into the optimized distributed power generation power injection power grid control model, and taking the distributed power generation power injection requirements and time constraints in the task as the input of the model;

[0053] Execute the distributed power generation power injection grid control model, perform simulation calculation according to the input task data and the parameters of the distributed power generation power injection grid control model, the distributed power generation power injection grid control model uses the current state of the grid, topology structure and real-time power data of the distributed power generation to calculate the distributed power generation power injection plan that meets the grid constraints and task requirements;

[0054] Extracting real-time distributed power generation power injection grid control parameters from the output of the distributed power generation power injection grid control model, the real-time distributed power generation power injection grid control parameters including the power adjustment amount of the distributed power generation, the adjustment parameters of the control strategy and the time point of power injection;

[0055] Before applying the grid control parameters of real-time distributed power generation power injection to the actual grid, the parameters of the grid simulation model of the optimized distributed power generation power injection are compared with the grid control parameters of the real-time distributed power generation power injection. If the grid constraints are met, the verification is successful. If the verification is successful, the grid control parameters of the real-time distributed power generation power injection are applied to the distributed power control system of the actual grid.

[0056] After applying the grid control parameters of real-time distributed power generation power injection, the state of the grid and the power injection of the distributed power generation are continuously monitored. If they do not meet expectations, the grid control parameters of real-time distributed power generation power injection are adjusted.

[0057] Beneficial effects of the present invention:

[0058] Through the global optimization capability of the genetic algorithm, the present invention can find the best power injection scheme to minimize losses, ensure voltage stability, and maximize the utilization of distributed power sources under the constraints of the power grid. This helps to improve the overall operating efficiency and stability of the power grid, reduce energy waste, and enhance the power grid's capacity to accommodate distributed power access.

[0059] The present invention simulates and evaluates the power injection scheme of distributed power sources through a power grid simulation model, providing a scientific basis for power grid dispatching. At the same time, combined with real-time power grid status data and task requirements, the present invention can quickly generate and adjust the power injection scheme, enhancing the flexibility of power grid dispatching. By optimizing the power injection scheme of distributed power sources, the present invention can maximize the use of the power generation capacity of distributed power sources and reduce the occurrence of wind and solar power abandonment. This helps to improve the utilization rate of renewable energy and promote the transformation and upgrading of the energy structure.

[0060] The present invention combines genetic algorithms, power grid simulation models and control models to construct an intelligent distributed power supply power injection optimization system. The present invention can automatically complete the formulation, verification and application of power injection schemes, thereby improving the intelligence level and automation level of the power grid. By optimizing the power injection scheme of distributed power sources, the present invention can reduce the loss and maintenance costs of the power grid. At the same time, improving the utilization rate of distributed power sources also helps to reduce dependence on traditional energy and procurement costs, thereby reducing the overall operating costs of the power grid.

[0061] The present invention takes into account the real-time status, topological structure and real-time power data of the power grid, so that the optimized power injection scheme can better adapt to the changes and fluctuations of the power grid, enhance the adaptability and resilience of the power grid, and improve the power grid's ability to respond to emergencies such as extreme weather and equipment failures.

[0062] In summary, the grid coordinated optimization operation method with distributed power generation power injection described in the present invention shows significant beneficial effects in improving grid efficiency and stability, enhancing the flexibility and accuracy of grid dispatching, promoting the efficient utilization of distributed power sources, improving the intelligence level of the grid, reducing grid operation costs, and enhancing the adaptability and resilience of the grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on the drawings without paying any creative labor.

[0064] Figure 1 A schematic flow chart of a method for coordinated optimization of power grid operation with distributed power generation injection provided by the present invention. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions provided by the embodiments of the present invention are described in detail below in conjunction with the drawings.

[0066] In order to better understand the purpose of the present invention, the present invention is described in further detail below.

[0067] The method for coordinated optimization operation of a power grid with distributed power generation power injection of the present invention comprises:

[0068] Step S101, obtaining real-time power data of distributed power sources, power grid status data, power grid topology data, distributed power source control strategy data, power grid dispatch center command data and historical operation data;

[0069] Real-time power data of distributed power sources: Through the interface with the monitoring system or supervisory control and data acquisition (SCADA) of distributed power sources, the current output power data of each distributed power source can be obtained in real time. Distributed power sources include but are not limited to solar power stations, wind power stations and energy storage systems.

[0070] Grid status data: Real-time grid status data is obtained from the grid monitoring system. These data include basic parameters such as voltage, current, frequency, power factor, etc. of each grid node, as well as grid load conditions and line status (such as whether it is overloaded or whether there is a fault).

[0071] Grid topology data: Obtain the topology information of the grid, including the connection relationship and parameters of the grid's nodes, lines, transformers, switchgear, etc. This information is the basis for understanding the grid structure, simulating and optimizing the grid.

[0072] Distributed power control strategy data: Obtain the control strategy information of distributed power, including power control strategy (such as constant power control, maximum power point tracking control, etc.), voltage control strategy and frequency control strategy. These strategy information reflects the operation mode and regulation capability of distributed power.

[0073] Power grid dispatch center command data: Obtain dispatch instructions for distributed power sources from the power grid dispatch center. These instructions include power dispatch instructions, start and stop instructions, reactive power compensation instructions, etc., which are used to coordinate power exchange between distributed power sources and the power grid.

[0074] Historical operation data: Obtain historical operation data from the historical database of the power grid and distributed power sources, including past power data, power grid status data, fault records, and maintenance records. Historical data plays an important role in analyzing the operation rules of the power grid and distributed power sources and improving the accuracy of the simulation model.

[0075] Through step S101, the method of the present invention can obtain the data of the power grid and the distributed power source comprehensively and in real time, providing a solid data foundation for subsequent data processing, simulation, optimization and control.

[0076] Step S102, obtaining natural environment data of the location where the power grid is located, performing cluster analysis on the natural environment data of the location where the power grid is located, obtaining a clustering result of the natural environment of the location where the power grid is located, establishing an association relationship between the clustering result of the natural environment of the location where the power grid is located and time information, and obtaining a knowledge base of the natural environment of the power grid;

[0077] First, the natural environment data of the location of the power grid is obtained, including but not limited to meteorological data (such as temperature, humidity, wind speed, wind direction, precipitation, air pressure, etc.), geographic data (such as terrain, altitude, soil type, etc.) and environmental data (such as air quality, solar radiation intensity, etc.).

[0078] The acquired natural environment data is preprocessed, including data cleaning, denoising, interpolation and other steps to improve data quality. Then, features are extracted from the preprocessed data, which are used to reflect the impact of the natural environment on the operation of the power grid. For example, meteorological features may include daily average temperature, maximum temperature, minimum temperature, average wind speed, precipitation, etc.

[0079] Clustering algorithms (such as K-means, hierarchical clustering, etc.) are used to perform cluster analysis on the extracted features, and similar natural environment data are clustered into clusters. The clustering results reflect the distribution laws and change patterns of the natural environment at different time or spatial scales.

[0080] The natural environment clustering results are associated with the corresponding time information (such as date, time period, season, etc.). This association helps to understand the changing trend of the natural environment over time and the possible impact of this change on the operation of the power grid.

[0081] The natural environment data, natural environment clustering results and their association with time information are integrated into a structured knowledge base. This knowledge base not only contains rich natural environment data, but also reveals the internal connections and change laws between these data, providing an important reference for subsequent power grid simulation, optimization and control.

[0082] Through step S102, the method of the present invention can establish a comprehensive and accurate knowledge base of the natural environment of the power grid, providing strong support for subsequent optimization of power grid operation.

[0083] Step S103, receiving a grid task for distributed power generation power injection, the grid task for distributed power generation power injection including basic task information, distributed power generation information and grid constraint data, matching the grid task for distributed power generation power injection in a grid natural environment knowledge base to obtain natural environment data corresponding to the grid task for distributed power generation power injection;

[0084] The grid task of receiving power injection from distributed power sources includes all the basic information required to perform the task, including task ID, task name, planned execution time, task priority, etc. In addition, the task also contains information about distributed power sources, such as the type of distributed power sources (such as solar power stations, wind power stations, energy storage systems, etc.), rated power, and access point location. The task also clarifies the constraint data of the power grid, such as the voltage level of the power grid, frequency stability requirements, power factor restrictions, etc. These constraints are key parameters to ensure the safe and stable operation of the power grid.

[0085] Matching in the power grid natural environment knowledge base, the system extracts information related to the task execution time from the basic task information, including the date and time period of the planned execution time. Using this time information, search and match in the previously constructed power grid natural environment knowledge base. The natural environment data in the knowledge base has been classified and organized according to time information (such as date, time period, season, etc.). Through the matching process, the system is able to find the natural environment data clustering results corresponding to the task execution time. The natural environment data clustering results include key natural environment parameters such as temperature, wind speed, precipitation, light intensity, etc. during the period.

[0086] When a matching natural environment data clustering result is found, the system extracts specific natural environment data from the knowledge base. The natural environment data will be used in subsequent power grid simulation, optimization and control processes to ensure that the distributed power generation power injection scheme can adapt to the natural environment conditions at the time and meet the constraints of the power grid.

[0087] Through step S103, the method of the present invention can ensure that the grid task of distributed power generation power injection can fully take into account the natural environmental conditions at the time when executing, thereby improving the safety and stability of grid operation.

[0088] Step S104, using the real-time power data of distributed power sources, the power grid state data, the power grid topology data, the distributed power source control strategy data, the power grid dispatch center command data and the historical operation data to train the random forest model, to obtain a power grid simulation model of distributed power source power injection, and to substitute the basic task information into the power grid simulation model of distributed power source power injection, to obtain a power grid simulation result of distributed power source power injection;

[0089] First, all acquired data are summarized, including real-time power data of distributed power sources, power grid status data, power grid topology data, distributed power source control strategy data, power grid dispatch center command data, and historical operation data. These data are preprocessed, including data cleaning (removing outliers, missing value processing, etc.), data transformation (such as normalization, standardization, etc.) and data dimensionality reduction (such as feature selection, principal component analysis, etc.) to improve data quality and reduce the complexity of model training.

[0090] Extract key features from the preprocessed data. These features may include real-time power output of distributed generation, voltage and current at each node of the grid, power flow of the line, control parameters of distributed generation, etc. The selection of features should be based on their impact on grid operation and their correlation with distributed generation power injection.

[0091] The dataset is divided into a training set and a test set. The training set is used to train the random forest model, while the test set is used to evaluate the performance of the model. When dividing, ensure that the training set and the test set have similar data distribution to avoid overfitting or underfitting.

[0092] Use the training set data to train the random forest model. Random forest is an ensemble learning method that improves the accuracy of the model by building multiple decision trees and combining their predictions. During the training process, you need to set the parameters of the random forest, such as the number of trees, maximum depth, minimum number of sample splits, etc. The choice of these parameters should be adjusted according to the characteristics of the specific problem and dataset.

[0093] Use the test set data to evaluate the trained random forest model, and measure the performance of the model by comparing the difference between the model prediction results and the actual results. According to the evaluation results, tune the model, such as adjusting parameters, adding features, etc., to improve the accuracy and generalization ability of the model.

[0094] Substitute the basic information of the task (such as task time, distributed power type, power data, etc.) into the trained random forest model. The model simulates the impact of distributed power injection on the power grid based on the input task information and the current state data of the power grid, and outputs the simulation results of the power grid with distributed power injection. The simulation results may include voltage changes at each node of the power grid, power flow of the line, power output data of the distributed power source, etc., which provide important basis for subsequent power grid optimization and control.

[0095] Through step S104, the method of the present invention can use machine learning technology to simulate and predict the impact of distributed power generation power injection on the power grid, thereby providing a scientific basis for the optimized operation of the power grid.

[0096] Step S105, compare the grid simulation result of distributed power generation power injection with the grid constraint data to obtain the comparison result of the grid simulation result of distributed power generation power injection; if the grid simulation result of distributed power generation power injection is outside the range of the grid constraint data, then optimize the grid simulation model of distributed power generation power injection using a genetic algorithm based on the real-time power data of the distributed power generation, the grid status data and the grid topology data to obtain an optimized grid simulation model of distributed power generation power injection; use the optimized grid simulation model of distributed power generation power injection as the grid control model of distributed power generation power injection; substitute the grid task of distributed power generation power injection into the grid control model of distributed power generation power injection to obtain real-time distributed power generation power injection grid control parameters.

[0097] In step S105, the method of the present invention conducts in-depth analysis and optimization of the power grid simulation results of distributed power generation power injection, and the specific steps are as follows:

[0098] The simulation results of the grid injected with distributed generation power are compared with the grid constraint data, which may include voltage level restrictions, frequency stability requirements, power factor range, line current carrying capacity restrictions, etc. The purpose of the comparison is to check whether the simulation results are within all grid constraints, that is, whether they meet the safe and stable operation requirements of the grid.

[0099] If the simulation results show that the grid state after the distributed generation power injection is within all constraints, then the simulation results are considered compliant and can be directly used for subsequent grid control. If the simulation results show that any one or more constraints are violated, then the simulation results are considered non-compliant and need to be optimized.

[0100] For non-compliant simulation results, a genetic algorithm is used to optimize the grid simulation model of distributed power generation power injection. Genetic algorithm is a heuristic search algorithm that solves optimization problems by simulating natural selection and genetic mechanisms. In the optimization process, the real-time power data of distributed power generation, grid status data, and grid topology data are used as inputs, and the degree of violation of grid constraints or achieving specific optimization goals (such as minimizing voltage deviation, minimizing line loss, etc.) are used as optimization goals. Through the iterative search of the genetic algorithm, a set of better distributed power generation power injection schemes are found so that the simulation results meet the grid constraints.

[0101] The optimized grid simulation model of distributed power generation power injection is used as a new grid control model of distributed power generation power injection. This control model can more accurately reflect the impact of distributed power generation power injection on the grid and meet the constraints of the grid.

[0102] Substitute the grid task of distributed power generation power injection into the optimized grid control model, and calculate the real-time grid control parameters of distributed power generation power injection based on the model's prediction results and the real-time status data of the grid. The control parameters may include the power output set value, reactive power compensation, voltage regulation, etc. of the distributed power generation, which will be used to guide the actual operation of the distributed power generation and the real-time control of the grid.

[0103] Through step S105, the method of the present invention can ensure that the power grid simulation results of distributed power generation power injection meet the constraints of the power grid, and obtain real-time power grid control parameters through optimization, providing strong guarantee for the safe and stable operation of the power grid.

[0104] Specifically, the grid coordinated optimization operation method for distributed power generation power injection according to the present invention, step S101 includes:

[0105] Through the interface with the monitoring system or data acquisition and monitoring control system of distributed power sources, the current output power data of each distributed power source is obtained in real time, and the real-time power data of the distributed power source is obtained. The distributed power source includes a solar power station, a wind power station and an energy storage system;

[0106] Acquire the real-time status data of the power grid from the monitoring system of the power grid. The real-time status data of the power grid includes the voltage, current, frequency, power factor of each node, the load condition of the power grid, and the line condition. The line condition includes the line condition overload and line fault, and obtain the power grid status data;

[0107] The topological structure information of the power grid includes the connection relationship and parameters of the nodes, lines, transformers, and switchgear of the power grid;

[0108] The control strategy information of the distributed power source includes power control strategy, voltage control strategy and frequency control strategy. The power control strategy includes constant power control and maximum power point tracking control.

[0109] Obtain dispatching instructions for distributed power sources from the power grid dispatching center, including power dispatching instructions, start and stop instructions, and reactive power compensation instructions, which are used to coordinate the power exchange between distributed power sources and the power grid and obtain the command data of the power grid dispatching center;

[0110] The historical operation data is obtained from the historical database of the power grid and distributed power sources. The historical operation data includes past power data, power grid status data, fault records, and maintenance records.

[0111] Through pre-established interfaces, it communicates with the monitoring system of distributed power sources (such as the monitoring system of solar power plants, wind power plants, and energy storage systems) or the data acquisition and monitoring control system (SCADA system). The current output power data of each distributed power source is obtained from these systems in real time. These data are the key to understanding the real-time power generation capacity of distributed power sources. The acquired data is sorted to form a real-time power data set of distributed power sources, which provides a basis for subsequent analysis and simulation.

[0112] The real-time status data of the power grid is extracted from the monitoring system of the power grid. These data include but are not limited to the voltage, current, frequency, power factor and other electrical parameters of each node. The monitoring line status includes whether the line is overloaded or faulty, etc. This information is crucial for evaluating the stability and safety of the power grid.

[0113] The topological structure information of the power grid is collected, including the connection relationship and parameters of electrical components such as nodes, lines, transformers, and switchgear in the power grid.

[0114] The control strategy information collected for distributed power sources includes power control strategy (such as constant power control, maximum power point tracking control, etc.), voltage control strategy and frequency control strategy. The control strategy determines the operation mode and response characteristics of distributed power sources in the power grid, which is crucial for optimizing the coordinated operation of distributed power sources and the power grid.

[0115] Communicate with the grid dispatch center to obtain dispatch instructions for distributed power sources. These instructions may include power dispatch instructions (specify the power generation of distributed power sources), start and stop instructions (control the start and stop of distributed power sources), reactive power compensation instructions (adjust the reactive power output of distributed power sources to improve the voltage quality of the grid), etc. Dispatching instructions for distributed power sources are an important basis for coordinating power exchange between distributed power sources and the grid and maintaining stable operation of the grid.

[0116] Extract historical operation data from the historical database of power grid and distributed power generation. These data include past power data, power grid status data, fault records, and maintenance records. Historical data is of great value for analyzing the operation rules of power grid and distributed power generation, predicting future trends, and optimizing operation strategies.

[0117] Specifically, the grid coordinated optimization operation method for distributed power generation power injection according to the present invention, step S102 includes:

[0118] Obtaining natural environment data of the location where the power grid is located. The natural environment data of the location where the power grid is located includes meteorological data, geographical data and environmental data. Meteorological data includes temperature, humidity, wind speed, wind direction, precipitation and air pressure. Geographic data includes topography, altitude and soil type. Environmental data includes air quality and solar radiation intensity.

[0119] Preprocess the natural environment data, extract features from the preprocessed natural environment data, and obtain meteorological features to reflect the impact of the natural environment on the operation of the power grid. The meteorological features include daily average temperature features, maximum temperature features, minimum temperature features, wind speed mean features, and precipitation features;

[0120] Use a clustering algorithm to perform cluster analysis on the extracted features to obtain natural environment clustering results, and establish an association relationship between the natural environment clustering results and the corresponding time information;

[0121] Integrate natural environment data, natural environment clustering results and their association with time information into a structured knowledge base.

[0122] Collect natural environment data of the location of the power grid, which covers three aspects: meteorology, geography and environment. Meteorological data specifically includes temperature, humidity, wind speed, wind direction, precipitation and air pressure, which directly reflect the local climate conditions and have a direct impact on the operation of the power grid and the power generation efficiency of distributed power sources.

[0123] Geographic data includes terrain, altitude, and soil type, which determine the difficulty of grid construction and the installation conditions of distributed power sources. Environmental data includes air quality and solar radiation intensity, which have an important impact on the aging of grid equipment and the power generation efficiency of distributed power sources (especially solar power stations).

[0124] The collected natural environment data is cleaned, outliers are removed, missing values ​​are filled, and necessary data conversion is performed to ensure the accuracy of the data. Features are extracted from the preprocessed data, and these features should be able to reflect the impact of the natural environment on the operation of the power grid. For example, the daily average temperature feature, the maximum temperature feature, and the minimum temperature feature can reflect the impact of temperature on the load of power grid equipment and the power generation efficiency of distributed power sources; the wind speed mean feature can reflect the impact of wind energy on the power generation efficiency of wind power stations; the precipitation feature can reflect the impact of precipitation on the safety of power grid equipment, etc.

[0125] Cluster analysis is performed on the extracted features using clustering algorithms (such as K-means, DBSCAN, etc.). The purpose of cluster analysis is to classify time periods or locations with similar natural environment characteristics into one category in order to better understand and predict the impact of the natural environment on power grid operation. Through cluster analysis, natural environment clustering results can be obtained, which reveal the power grid operation mode and distributed generation characteristics under different natural environment characteristics.

[0126] Establish an association between the natural environment clustering results and the corresponding time information (such as date, season, etc.). This association helps to analyze the impact of changes in the natural environment over time on the operation of the power grid and the power generation efficiency of distributed power sources. For example, the impact of changes in meteorological characteristics such as temperature and wind speed in different seasons on the power grid load and the power generation efficiency of distributed power sources can be analyzed, providing a scientific basis for the dispatch and operation of the power grid.

[0127] Integrate natural environment data, natural environment clustering results and their association with time information into a structured knowledge base. This knowledge base should contain rich natural environment information and power grid operation information, and support data query, analysis and mining. By building a knowledge base, we can achieve an in-depth understanding and efficient use of the relationship between the natural environment and power grid operation, and provide strong support for the coordinated optimization operation of the power grid.

[0128] Specifically, the grid coordinated optimization operation method for distributed power generation power injection according to the present invention, step S103 includes:

[0129] Receive the grid task injected by the distributed power generation, parse the received grid task data, and extract the basic task information, distributed power generation information and grid constraint data;

[0130] The basic information of the task includes the task ID, task name, planned execution time and task priority;

[0131] Distributed power information includes the type, rated power and access point location of distributed power;

[0132] Grid constraint data includes the voltage level, frequency stability requirements, and power factor limits of the grid;

[0133] Extracting task execution time information from the task basic information, the task execution time information includes the date and time period of the planned execution time, using the task execution time information to search for natural environment data clustering results corresponding to the task execution time in the power grid natural environment knowledge base, the matching process involves searching for natural environment data corresponding to a specific date, time period or season;

[0134] Extracting natural environment data matching the task execution time from the power grid natural environment knowledge base, the natural environment data including temperature, wind speed, precipitation, light information and wind speed information during the period;

[0135] The acquired natural environment data is integrated with the grid mission data injected by distributed generation power, and the integrated data is used for subsequent grid simulation, optimization and control processes.

[0136] Receive grid tasks for distributed power generation power injection from superior dispatchers or users. Parse the received grid task data and extract key information, including basic task information, distributed power generation information, and grid constraint data.

[0137] Basic task information: includes task ID (unique identifier), task name, planned execution time (specific date and time period), and task priority.

[0138] Distributed power generation information: clearly specify the type of distributed power generation (such as solar power stations, wind power stations, energy storage systems, etc.), rated power, and specific location of access to the grid.

[0139] Grid constraint data: Lists the constraints that the grid must comply with for operation, such as voltage levels, frequency stability requirements, and power factor limitations.

[0140] Extract task execution time information: Extract the time information related to the task execution from the basic task information, that is, the specific date and time period of the planned execution time.

[0141] Matching natural environment data: Use the extracted task execution time information to search in the previously constructed power grid natural environment knowledge base.

[0142] The matching process involves finding the corresponding natural environment data cluster results according to a specific date, time period or season. This includes temperature, wind speed, precipitation, light intensity (especially important for solar power plants) and wind speed information (especially important for wind power plants) during that period.

[0143] Extract and integrate natural environment data: Extract natural environment data that accurately matches the task execution time from the power grid natural environment knowledge base. Integrate these natural environment data with the power grid task data injected by distributed generation power. The integrated data set not only contains detailed information about the task, but also incorporates the natural environment conditions that may be encountered when executing the task.

[0144] Prepare for subsequent processing: The integrated data is used for subsequent grid simulation, optimization and control processes. In the simulation stage, these data help build a grid model that is closer to reality; in the optimization stage, they support the search for solutions that maximize the utilization of distributed power sources and grid efficiency while meeting grid constraints; in the control stage, they are used to adjust the output of distributed power sources in real time to cope with changes in the natural environment and changes in grid demand.

[0145] Specifically, the grid coordinated optimization operation method for distributed power generation power injection according to the present invention, step S104 includes:

[0146] Summarize the real-time power data of distributed power sources, power grid status data, power grid topology data, distributed power source control strategy data, power grid dispatch center command data and historical operation data;

[0147] Preprocess the data and extract features from the preprocessed data to obtain the real-time power of the distributed power source, the voltage of each node in the power grid, the current of each node in the power grid, the power flow of the line, and the control parameters of the distributed power source;

[0148] The data set is divided into a training set and a test set. The training set is used to train the random forest model, and the test set is used to evaluate the performance of the model.

[0149] Use the training set data to train the random forest model, receive and set the parameters of the random forest, the parameters of the random forest include the number of trees, the maximum depth, and the minimum number of sample splits, run the random forest algorithm, obtain the power grid simulation model of distributed power generation injection, substitute the basic information of the power grid task of distributed power generation injection into the power grid simulation model of distributed power generation injection, the basic information of the power grid task includes the task time, the type of distributed power generation and the power data, run the power grid simulation model of distributed power generation injection, and obtain the power grid simulation results of distributed power generation injection, which include the voltage changes of each node of the power grid, the power flow of the line and the power output data of the distributed power generation.

[0150] Summarize all the data collected in the previous steps, including real-time power data of distributed power sources, grid status data, grid topology data, distributed power source control strategy data, grid dispatch center command data, and historical operation data.

[0151] The aggregated data is preprocessed, including data cleaning (removing outliers, filling missing values, etc.), data conversion (such as standardization, normalization, etc.), and data screening (selecting features useful for model training). Features are extracted from the preprocessed data, which should be able to reflect the impact of distributed power injection on the power grid. For example, the real-time power of distributed power, the voltage and current of each node of the power grid, the power flow of the line, the control parameters of distributed power, etc.

[0152] The dataset after preprocessing and feature extraction is divided into a training set and a test set. The training set is used to train the random forest model, while the test set is used to evaluate the performance of the model, such as accuracy and recall. Use the training set data to train the random forest model. Random forest is an ensemble learning method that improves the accuracy and stability of the model by building multiple decision trees and combining their prediction results. During the training process, it is necessary to receive and set the parameters of the random forest, such as the number of trees, maximum depth, minimum number of sample splits, etc. These parameters affect the complexity and performance of the model. Run the random forest algorithm to obtain a grid simulation model with distributed power generation injection.

[0153] Substitute the basic information of the power grid task of distributed power generation into the trained power grid simulation model. This information includes task time, distributed power type, and power data. Run the power grid simulation model of distributed power generation power injection to simulate the impact of distributed power generation power injection on the power grid. The simulation results obtained include voltage changes at each node of the power grid, power flow of the line, and power output data of the distributed power generation. These results can be used to evaluate the impact of distributed power generation power injection on the stability and efficiency of the power grid, and provide decision support for subsequent power grid optimization and control.

[0154] Through the implementation of step S104, the present invention can use the random forest model to accurately simulate and predict the impact of distributed power generation power injection on the power grid, providing a powerful tool and method for the coordinated optimization operation of the power grid. This data-driven method can improve the flexibility of power grid scheduling and help achieve efficient utilization of distributed power sources and stable operation of the power grid.

[0155] Specifically, the grid coordinated optimization operation method for distributed power generation power injection according to the present invention, step S105 includes:

[0156] Determine the optimization target, based on the actual needs and constraints of the power grid, to determine the optimization target, which includes minimizing the loss of the power grid, stabilizing the voltage, and maximizing the utilization of distributed power sources;

[0157] Receive and set basic parameters of the genetic algorithm, including population size, genetic generation, crossover probability, and mutation probability;

[0158] The power injection scheme of the distributed generation is represented as an individual in the genetic algorithm by using real number coding, and the individual includes the power setting value or control parameter of the distributed generation, and the individual is taken as the object of optimization of the genetic algorithm;

[0159] Initialize the population. According to the set population size, randomly generate a group of initial individuals to form the initial population. Each individual represents a distributed power injection scheme.

[0160] Using the real-time power data of distributed generation, grid status data and grid topology data, the fitness of each individual is evaluated through the grid simulation model. The fitness function should be designed according to the optimization goal.

[0161] According to the fitness evaluation results, individuals with higher fitness are selected as parents for crossover and mutation operations;

[0162] Perform crossover operation on the selected parent individuals to generate new offspring individuals;

[0163] The crossover operation is performed by single-point crossover, double-point crossover or uniform crossover;

[0164] Perform mutation operations on offspring individuals by changing the gene values ​​in the individuals or introducing new genes;

[0165] Update the population, replace some or all of the parent individuals with the generated offspring individuals, update the population, and determine whether to continue iterating based on the set genetic generation or convergence conditions;

[0166] If the set genetic generation number is reached or the fitness meets the convergence condition, the iteration is stopped. After iterative optimization of the genetic algorithm, the individual with the highest fitness is obtained, that is, the optimized distributed power generation power injection scheme, which is used as the parameter of the optimized distributed power generation power injection power grid simulation model.

[0167] In step S105, the grid coordinated optimization operation method for distributed power generation power injection of the present invention uses a genetic algorithm to optimize the power injection scheme of the distributed power generation. The following is a detailed process of this step:

[0168] Determine the optimization objectives and clarify the optimization objectives based on the actual needs and constraints of the power grid. These objectives usually include minimizing the losses of the power grid, ensuring voltage stability, and maximizing the utilization of distributed generation. These objectives are interrelated and need to be weighed during the optimization process.

[0169] Set genetic algorithm parameters: Receive and set the basic parameters of the genetic algorithm, including population size (i.e. the number of individuals participating in the optimization at the same time), genetic generation (i.e. the number of optimization iterations), crossover probability (i.e. the probability of crossover between individuals), and mutation probability (i.e. the probability of mutation in an individual). These parameters will affect the performance and convergence speed of the genetic algorithm.

[0170] Encoded individuals: The power injection scheme of distributed generation is represented as individuals in the genetic algorithm using real number encoding. Each individual contains a set of power setpoints or control parameters of distributed generation, which are the variables that can be optimized.

[0171] Initialize the population: According to the set population size, a group of initial individuals are randomly generated to form the initial population. Each individual represents a possible distributed power injection scheme.

[0172] Evaluate fitness: Use the real-time power data of distributed power sources, grid status data, and grid topology data to evaluate the fitness of each individual through the previously trained grid simulation model. The fitness function should be designed according to the optimization goal. For example, the fitness function can be defined by comprehensively considering factors such as grid loss, voltage stability, and distributed power utilization.

[0173] Select parent generation: According to the fitness evaluation results, select individuals with higher fitness as parents for subsequent crossover and mutation operations. The selection strategy can adopt roulette wheel selection, tournament selection and other methods.

[0174] Crossover operation:

[0175] Perform a crossover operation on the selected parent individuals to generate new offspring individuals. The crossover operation can be performed by single-point crossover, double-point crossover or uniform crossover to simulate the recombination process of genes in biological evolution.

[0176] Mutation operation: Perform mutation operations on offspring individuals to increase the diversity of the population and explore new solution spaces. Mutation operations can be achieved by changing the gene values ​​in individuals or introducing new genes.

[0177] Update population: Replace some or all of the parent individuals with the generated offspring individuals to update the population. This step simulates the natural selection process in biological evolution. Determine whether to continue iteration based on the set genetic generation or convergence condition. If the set genetic generation is reached or the fitness meets the convergence condition, the iteration stops.

[0178] Obtain optimization results: After iterative optimization of the genetic algorithm, the individual with the highest fitness is obtained, that is, the optimized distributed power injection scheme. This scheme is the best solution to achieve the optimization goals (such as minimizing grid losses, ensuring voltage stability, and maximizing distributed power utilization) under the conditions of grid constraints.

[0179] The optimization scheme is used as the parameter of the power grid simulation model for optimized distributed generation power injection for subsequent power grid operation simulation, optimization and control.

[0180] By implementing step S105, the present invention can use the genetic algorithm to globally optimize the power injection scheme of the distributed power source and find the best solution for achieving the optimization goal while satisfying the grid constraints.

[0181] Specifically, the grid coordinated optimization operation method for distributed power generation power injection according to the present invention, step S105 includes:

[0182] The power grid simulation model of distributed power generation power injection is optimized by genetic algorithm, and the optimized model is obtained, and the parameters of the optimized power grid simulation model of distributed power generation power injection are set as the initial parameters of the power grid control model of distributed power generation power injection;

[0183] Receive the grid task of distributed generation power injection from the grid dispatch center. The grid task should include basic information of the task, such as task time, distributed generation type and expected power injection amount;

[0184] Preprocessing the received power grid task data, substituting the preprocessed power grid task data into the optimized distributed power generation power injection power grid control model, and taking the distributed power generation power injection requirements and time constraints in the task as the input of the model;

[0185] Execute the distributed power generation power injection grid control model, perform simulation calculation according to the input task data and the parameters of the distributed power generation power injection grid control model, the distributed power generation power injection grid control model uses the current state of the grid, topology structure and real-time power data of the distributed power generation to calculate the distributed power generation power injection plan that meets the grid constraints and task requirements;

[0186] Extracting real-time distributed power generation power injection grid control parameters from the output of the distributed power generation power injection grid control model, the real-time distributed power generation power injection grid control parameters including the power adjustment amount of the distributed power generation, the adjustment parameters of the control strategy and the time point of power injection;

[0187] Before applying the grid control parameters of real-time distributed power generation power injection to the actual grid, the parameters of the grid simulation model of the optimized distributed power generation power injection are compared with the grid control parameters of the real-time distributed power generation power injection. If the grid constraints are met, the verification is successful. If the verification is successful, the grid control parameters of the real-time distributed power generation power injection are applied to the distributed power control system of the actual grid.

[0188] After applying the grid control parameters of real-time distributed power generation power injection, the state of the grid and the power injection of the distributed power generation are continuously monitored. If they do not meet expectations, the grid control parameters of real-time distributed power generation power injection are adjusted.

[0189] In step S105, the present invention describes in detail how to formulate and apply actual grid control parameters using the grid simulation model of distributed power generation power injection optimized by genetic algorithm. The following is a detailed process of this step:

[0190] The grid simulation model of distributed power generation power injection is optimized by genetic algorithm to find the optimal or approximately optimal distributed power generation power injection scheme. The optimized model parameters are set as the initial parameters of the grid control model of distributed power generation power injection. These parameters include the power setting value of the distributed power generation, control strategy parameters, etc., which will serve as the basis for subsequent control decisions.

[0191] Receive the grid task of distributed generation power injection from the grid dispatch center. The task information should contain the basic elements of the task, such as the task execution time, the type of distributed generation involved, and the expected power injection amount. Preprocess the received grid task data to ensure the accuracy and consistency of the data. This may include steps such as data cleaning, format conversion and verification. Substitute the preprocessed task data into the optimized grid control model for distributed generation power injection.

[0192] Execute the grid control model for distributed power generation power injection, and perform simulation calculations based on the input mission data and model parameters. The model should comprehensively consider the current state of the grid, the topology, the real-time power data of the distributed power generation, and the mission requirements, and calculate the distributed power generation power injection scheme that meets the grid constraints and mission requirements. Extract the grid control parameters for real-time distributed power generation power injection from the model output. These parameters may include the power adjustment amount of the distributed power generation, the adjustment parameters of the control strategy, and the specific time point of power injection.

[0193] Before applying the control parameters to the actual power grid, the rationality and feasibility of the control parameters are verified by comparing the optimized power grid simulation model parameters with the real-time control parameters. If the verification is successful, that is, the control parameters meet the power grid constraints, they can be applied to the distributed power control system of the actual power grid. The successfully verified power grid control parameters for real-time distributed power generation power injection are applied to the actual power grid. After applying the control parameters, the state of the power grid and the power injection of the distributed power generation are continuously monitored to ensure that the power grid operates stably and the distributed power generation injects power as expected.

[0194] If the monitoring results show that the grid state or the distributed power generation power injection situation does not meet expectations, the grid control parameters of the real-time distributed power generation power injection should be adjusted in time. If necessary, the grid control model can be re-run or other optimization methods can be used to find more suitable control parameters, and then verified and applied again.

[0195] By implementing this step S105, the present invention can ensure that the grid control parameters for distributed power generation power injection are based on the results of the optimization model and meet the actual operation requirements of the grid.

[0196] The grid coordinated optimization operation method for distributed power generation power injection provided by the present invention effectively solves the problem that the output power of distributed power generation is affected by factors such as weather and operating conditions, showing strong volatility and intermittency, which increases the difficulty of grid frequency and voltage control and makes it difficult for traditional deterministic planning to adapt to the construction of new power systems through the following technical means:

[0197] The present invention acquires the power data of distributed power sources, power grid status data, power grid topology data, distributed power source control strategy data, power grid dispatch center instruction data and historical operation data in real time. The real-time acquisition of these data provides a basis for subsequent analysis and optimization.

[0198] Obtain the natural environment data (including meteorological, geographical, and environmental data) of the power grid location, perform cluster analysis, associate the clustering results with time information, and build a knowledge base of the power grid's natural environment. This step enables the system to predict the impact of changes in the natural environment on the operation of the power grid, so that adjustments can be made in advance.

[0199] After receiving the power grid task, the natural environment knowledge base of the power grid is matched to obtain the natural environment data corresponding to the task execution time. After these data are integrated with the power grid task data, they provide more comprehensive input for subsequent power grid simulation and optimization. The random forest model is trained using real-time and historical data to obtain a power grid simulation model. By substituting the basic information of the task, the state of the power grid after the distributed power generation is injected is simulated. The random forest model can handle high-dimensional data and complex nonlinear relationships, and improve the accuracy of the simulation.

[0200] The simulation results are compared with the grid constraint data. If the constraint conditions are not met, the simulation model is optimized using a genetic algorithm. The genetic algorithm can handle the complex optimization problem of distributed power injection by iteratively searching for the optimal solution, and improve the operation efficiency and stability of the grid. The simulation model optimized by the genetic algorithm is used as a grid control model, and outputs real-time control parameters after receiving the grid task. These parameters include the power adjustment amount of the distributed power source, the adjustment parameters of the control strategy, and the time point of power injection, which are directly applied to the distributed power control system of the actual grid. After applying the control parameters, the grid status and the power injection of the distributed power source are continuously monitored. If it does not meet expectations, the control parameters are adjusted in time to ensure the stable operation of the grid.

[0201] Through the above-mentioned technical means, the present invention realizes the precise control and optimization of the power injection of distributed power sources, effectively solves the grid control problem caused by the volatility and intermittency of the output power of distributed power sources, improves the operation efficiency, stability and safety of the grid, and adapts to the construction needs of new power systems.

Claims

1. A method for coordinated optimization operation of a power grid with distributed power generation injection, characterized in that: include: Step S101, obtaining real-time power data of distributed power sources, power grid status data, power grid topology data, distributed power source control strategy data, power grid dispatch center command data and historical operation data; Step S102, obtaining natural environment data of the location where the power grid is located, performing cluster analysis on the natural environment data of the location where the power grid is located, obtaining a clustering result of the natural environment of the location where the power grid is located, establishing an association relationship between the clustering result of the natural environment of the location where the power grid is located and time information, and obtaining a knowledge base of the natural environment of the power grid; Step S103, receiving a grid task for distributed power generation power injection, the grid task for distributed power generation power injection including basic task information, distributed power generation information and grid constraint data, matching the grid task for distributed power generation power injection in a grid natural environment knowledge base to obtain natural environment data corresponding to the grid task for distributed power generation power injection; Step S104, using the real-time power data of distributed power sources, the power grid state data, the power grid topology data, the distributed power source control strategy data, the power grid dispatch center command data and the historical operation data to train the random forest model, to obtain a power grid simulation model of distributed power source power injection, and to substitute the basic task information into the power grid simulation model of distributed power source power injection, to obtain a power grid simulation result of distributed power source power injection; Step S105, compare the grid simulation result of distributed power generation power injection with the grid constraint data to obtain the comparison result of the grid simulation result of distributed power generation power injection; if the grid simulation result of distributed power generation power injection is outside the range of the grid constraint data, then optimize the grid simulation model of distributed power generation power injection using a genetic algorithm based on the real-time power data of the distributed power generation, the grid status data and the grid topology data to obtain an optimized grid simulation model of distributed power generation power injection; use the optimized grid simulation model of distributed power generation power injection as the grid control model of distributed power generation power injection; substitute the grid task of distributed power generation power injection into the grid control model of distributed power generation power injection to obtain real-time distributed power generation power injection grid control parameters.

2. The method for coordinated optimization of power grid operation with distributed power generation injection according to claim 1, characterized in that: The step S101 includes: Through the interface with the monitoring system or data acquisition and monitoring control system of distributed power sources, the current output power data of each distributed power source is obtained in real time, and the real-time power data of the distributed power source is obtained. The distributed power source includes a solar power station, a wind power station and an energy storage system; Acquire the real-time status data of the power grid from the monitoring system of the power grid. The real-time status data of the power grid includes the voltage, current, frequency, power factor of each node, the load condition of the power grid, and the line condition. The line condition includes the line condition overload and line fault, and obtain the power grid status data; The topological structure information of the power grid includes the connection relationship and parameters of the nodes, lines, transformers, and switchgear of the power grid; The control strategy information of the distributed power source includes power control strategy, voltage control strategy and frequency control strategy. The power control strategy includes constant power control and maximum power point tracking control. Obtain dispatching instructions for distributed power sources from the power grid dispatching center, including power dispatching instructions, start and stop instructions, and reactive power compensation instructions, which are used to coordinate the power exchange between distributed power sources and the power grid and obtain the command data of the power grid dispatching center; The historical operation data is obtained from the historical database of the power grid and distributed power sources. The historical operation data includes past power data, power grid status data, fault records, and maintenance records.

3. The method for coordinated optimization operation of a power grid with distributed power generation injection according to claim 1, characterized in that: The step S102 includes: Obtaining natural environment data of the location where the power grid is located. The natural environment data of the location where the power grid is located includes meteorological data, geographical data and environmental data. Meteorological data includes temperature, humidity, wind speed, wind direction, precipitation and air pressure. Geographic data includes topography, altitude and soil type. Environmental data includes air quality and solar radiation intensity. Preprocess the natural environment data, extract features from the preprocessed natural environment data, and obtain meteorological features to reflect the impact of the natural environment on the operation of the power grid. The meteorological features include daily average temperature features, maximum temperature features, minimum temperature features, wind speed mean features, and precipitation features; Use a clustering algorithm to perform cluster analysis on the extracted features to obtain natural environment clustering results, and establish an association relationship between the natural environment clustering results and the corresponding time information; Integrate natural environment data, natural environment clustering results and their association with time information into a structured knowledge base.

4. The method for coordinated optimization of power grid operation with distributed power generation injection according to claim 1, characterized in that: The step S103 includes: Receive the grid task injected by the distributed power generation, parse the received grid task data, and extract the basic task information, distributed power generation information and grid constraint data; The basic information of the task includes the task ID, task name, planned execution time and task priority; Distributed power information includes the type, rated power and access point location of distributed power; Grid constraint data includes the voltage level, frequency stability requirements, and power factor limits of the grid; Extracting task execution time information from the task basic information, the task execution time information includes the date and time period of the planned execution time, using the task execution time information to search for natural environment data clustering results corresponding to the task execution time in the power grid natural environment knowledge base, the matching process involves searching for natural environment data corresponding to a specific date, time period or season; Extracting natural environment data matching the task execution time from the power grid natural environment knowledge base, the natural environment data including temperature, wind speed, precipitation, light information and wind speed information during the period; The acquired natural environment data is integrated with the grid mission data injected by distributed generation power, and the integrated data is used for subsequent grid simulation, optimization and control processes.

5. The method for coordinated optimization operation of a power grid with distributed power generation injection as claimed in claim 1, characterized in that: The step S104 includes: Summarize the real-time power data of distributed power sources, power grid status data, power grid topology data, distributed power source control strategy data, power grid dispatch center command data and historical operation data; Preprocess the data and extract features from the preprocessed data to obtain the real-time power of the distributed power source, the voltage of each node in the power grid, the current of each node in the power grid, the power flow of the line, and the control parameters of the distributed power source; The data set is divided into a training set and a test set. The training set is used to train the random forest model, and the test set is used to evaluate the performance of the model. Use the training set data to train the random forest model, receive and set the parameters of the random forest, the parameters of the random forest include the number of trees, the maximum depth, and the minimum number of sample splits, run the random forest algorithm, obtain the power grid simulation model of distributed power generation injection, substitute the basic information of the power grid task of distributed power generation injection into the power grid simulation model of distributed power generation injection, the basic information of the power grid task includes the task time, the type of distributed power generation and the power data, run the power grid simulation model of distributed power generation injection, and obtain the power grid simulation results of distributed power generation injection, which include the voltage changes of each node of the power grid, the power flow of the line and the power output data of the distributed power generation.

6. The method for coordinated optimization operation of a power grid with distributed power generation injection as claimed in claim 1, characterized in that: The step S105 includes: Determine the optimization target, based on the actual needs and constraints of the power grid, to determine the optimization target, which includes minimizing the loss of the power grid, stabilizing the voltage, and maximizing the utilization of distributed power sources; Receive and set basic parameters of the genetic algorithm, including population size, genetic generation, crossover probability, and mutation probability; The power injection scheme of the distributed generation is represented as an individual in the genetic algorithm by using real number coding, and the individual includes the power setting value or control parameter of the distributed generation, and the individual is taken as the object of optimization of the genetic algorithm; Initialize the population. According to the set population size, randomly generate a group of initial individuals to form the initial population. Each individual represents a distributed power injection scheme. Using the real-time power data of distributed generation, grid status data and grid topology data, the fitness of each individual is evaluated through the grid simulation model. The fitness function should be designed according to the optimization goal. According to the fitness evaluation results, individuals with higher fitness are selected as parents for crossover and mutation operations; Perform crossover operation on the selected parent individuals to generate new offspring individuals; The crossover operation is performed by single-point crossover, double-point crossover or uniform crossover; Perform mutation operations on offspring individuals by changing the gene values ​​in the individuals or introducing new genes; Update the population, replace some or all of the parent individuals with the generated offspring individuals, update the population, and determine whether to continue iterating based on the set genetic generation or convergence conditions; If the set genetic generation number is reached or the fitness meets the convergence condition, the iteration is stopped. After iterative optimization of the genetic algorithm, the individual with the highest fitness is obtained, that is, the optimized distributed power generation power injection scheme, which is used as the parameter of the optimized distributed power generation power injection power grid simulation model.

7. The method for coordinated optimization operation of a power grid with distributed power generation injection as claimed in claim 6, characterized in that: The step S105 includes: The power grid simulation model of distributed power generation power injection is optimized by genetic algorithm, and the optimized model is obtained, and the parameters of the optimized power grid simulation model of distributed power generation power injection are set as the initial parameters of the power grid control model of distributed power generation power injection; Receive the grid task of distributed generation power injection from the grid dispatch center. The grid task should include basic information of the task, such as task time, distributed generation type and expected power injection amount; Preprocessing the received power grid task data, substituting the preprocessed power grid task data into the optimized distributed power generation power injection power grid control model, and taking the distributed power generation power injection requirements and time constraints in the task as the input of the model; Execute the distributed power generation power injection grid control model, perform simulation calculation according to the input task data and the parameters of the distributed power generation power injection grid control model, the distributed power generation power injection grid control model uses the current state of the grid, topology structure and real-time power data of the distributed power generation to calculate the distributed power generation power injection plan that meets the grid constraints and task requirements; Extracting real-time distributed power generation power injection grid control parameters from the output of the distributed power generation power injection grid control model, the real-time distributed power generation power injection grid control parameters including the power adjustment amount of the distributed power generation, the adjustment parameters of the control strategy and the time point of power injection; Before applying the grid control parameters of real-time distributed power generation power injection to the actual grid, the parameters of the grid simulation model of the optimized distributed power generation power injection are compared with the grid control parameters of the real-time distributed power generation power injection. If the grid constraints are met, the verification is successful. If the verification is successful, the grid control parameters of the real-time distributed power generation power injection are applied to the distributed power control system of the actual grid. After applying the grid control parameters of real-time distributed power generation power injection, the state of the grid and the power injection of the distributed power generation are continuously monitored. If they do not meet expectations, the grid control parameters of real-time distributed power generation power injection are adjusted.

Images