Multi-target reactive compensation planning method, device and equipment and storage medium

Abstract

The invention provides a multi-target reactive power compensation planning method, device and equipment and a storage medium, and the method comprises the steps: employing K-means clustering new energy output data as a typical scene, calculating the L index of each scene node and the sensitivity of the L index, and selecting the first 10 high-frequency nodes of the sensitivity as compensation points; and based on L index improvement and network loss minimization, performing multi-objective optimization solution to obtain optimal configuration of each scene. And determining a compensation node type and a constraint condition, simulating and verifying a compensation effect, and if no-difference compensation is achieved, adjusting parameters and performing re-optimization. And after verification is qualified, the scheme is applied to practice, and efficient reactive compensation is realized. According to the method, the reasonability is enhanced by verifying the feedback optimization scheme.

Description

Technical Field

[0001] The present application relates to the field of reactive power compensation for new energy power grids, and in particular to a multi-objective reactive power compensation planning method, device, equipment and storage medium. Background Art

[0002] The output of new energy sources (such as wind power and solar energy) has significant uncertainty and volatility. This characteristic makes it difficult to predict and control the voltage stability, power quality and network loss level of the power system. At the same time, the modern power system has a complex structure, contains a large number of nodes and lines, and its operating status is affected by many factors (such as load changes, equipment failures, etc.). Therefore, when planning reactive power compensation, it is necessary to consider the operating characteristics of the power grid in various scenarios to ensure that the voltage can be maintained stable and the network loss can be reduced under various circumstances.

[0003] Traditional methods often select compensation nodes based on experience or simple rules, which makes it difficult to achieve the optimal compensation effect in a complex and changing power grid environment. In addition, different types of reactive power sources (such as capacitors, phase regulators, etc.) have different technical characteristics and economic costs. How to reasonably select and configure reactive power sources is also a major challenge. Summary of the invention

[0004] The purpose of this application is to overcome the problems existing in the above-mentioned prior art and to provide a multi-objective reactive power compensation planning method, device, equipment and storage medium.

[0005] The present application provides a multi-objective reactive power compensation planning method, including:

[0006] The K-means algorithm is used to cluster the renewable energy output data into multiple typical scenarios;

[0007] Calculate the L index and L index sensitivity of each node in each typical scenario;

[0008] Select the top 10 candidate nodes with the initial sensitivity value of the L index in each typical scenario;

[0009] The top five candidate nodes with the highest appearance frequency among all the candidate nodes are selected as reactive power compensation nodes;

[0010] The objective function is set according to the L index improvement of the top 10 nodes in each scenario and the minimization of the system network loss, and a multi-objective optimization solution is used to obtain the optimal reactive power configuration solution for each typical scenario;

[0011] Determining the reactive source configuration type of the reactive compensation node and system constraints;

[0012] According to the optimal reactive power configuration scheme, reactive power source configuration type and the constraint conditions, simulated compensation is performed for each of the typical scenarios to determine whether the requirement of zero-difference compensation is met;

[0013] If the requirement of no-difference compensation is not met, identify the nodes and their reactive power sources that do not meet the requirement of no-difference compensation;

[0014] Adjust the parameters of nodes and their reactive sources that do not meet the requirements of zero-difference compensation and perform multi-objective optimization again;

[0015] If the requirement of zero-difference compensation is met, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation. Optionally, the L index sensitivity LQSIj is calculated as follows:

[0016] LQSIj = (LL (ΔQ)) / ΔQ;

[0017] Where L is the L index value before reactive power optimization, L(ΔQ) is the L index value after reactive power compensation, and ΔQ is the reactive power compensation amount.

[0018] Optionally, the reactive source configuration type includes one or more of capacitors, phase regulators, SVCs, and SVGs.

[0019] Optionally, if the requirement of no-difference compensation is met, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation, including:

[0020] Monitor the static voltage stability of the power system, evaluate the reactive power compensation effect, and adjust the optimization strategy based on the evaluation results.

[0021] Optionally, the K-means algorithm is used to cluster the renewable energy output data into multiple typical scenarios, including:

[0022] According to the time series characteristics of renewable energy output, load changes and weather factors, the K-means algorithm is used to cluster the renewable energy output data.

[0023] Optionally, after selecting the top 10 candidate nodes of the initial sensitivity value of the L index in each typical scenario, it includes:

[0024] A voltage stability assessment is performed on the candidate nodes to ensure that the selected candidate nodes can improve the overall voltage stability of the system after reactive power compensation.

[0025] Optionally, it also includes:

[0026] A system model is established in the power system simulation software, including the renewable energy power generation model, load model, and transmission network model, to more accurately simulate the operation of the actual power system.

[0027] The present application also provides a multi-objective reactive power compensation planning device, comprising:

[0028] The scenario module uses the K-means algorithm to cluster the renewable energy output data into multiple typical scenarios;

[0029] A calculation module, calculating the L index and L index sensitivity of each node in each typical scenario;

[0030] A selection module selects the top 10 candidate nodes with the initial sensitivity value of the L indicator in each typical scenario;

[0031] A node module, taking the top 5 candidate nodes with the highest appearance frequency among all the candidate nodes as reactive power compensation nodes;

[0032] The optimization module sets the objective function according to the L index improvement of the top 10 nodes in each scenario and the minimization of the system network loss, and uses multi-objective optimization to obtain the optimal reactive power configuration solution under each typical scenario;

[0033] A condition module, determining the reactive source configuration type of the reactive compensation node and system constraints;

[0034] A verification module, which performs simulated compensation for each of the typical scenarios according to the optimal reactive power configuration scheme, reactive power source configuration type and the constraint conditions, and determines whether the requirement of zero-difference compensation is met;

[0035] The identification module identifies the nodes and their reactive power sources that do not meet the requirements of the differential compensation if the requirements of the differential compensation are not met;

[0036] Feedback module, adjusts the parameters of nodes and their reactive sources that do not meet the requirements of zero-difference compensation and performs multi-objective optimization again;

[0037] The execution module, if the requirements of no-difference compensation are met, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation.

[0038] The present application also provides a multi-objective reactive power compensation planning device, comprising:

[0039] Memory;

[0040] A processor is used to retrieve a computer executable program of the above-mentioned multi-objective reactive power compensation planning method from the memory, and execute: applying the K-means algorithm to cluster the renewable energy output data into multiple typical scenarios; calculating the L index and L index sensitivity of each node in each of the typical scenarios; selecting the top 10 candidate nodes with the initial value of the L index sensitivity in each of the typical scenarios; selecting the top 5 candidate nodes with the highest frequency of occurrence among all the candidate nodes as reactive power compensation nodes; setting the objective function according to the L index improvement of the top 10 nodes in each scenario and the minimization of the system network loss, and using multi-objective optimization to solve, Obtain the optimal reactive power configuration scheme under each of the typical scenarios; determine the reactive source configuration type of the reactive compensation node and the system constraints; perform simulated compensation for each of the typical scenarios according to the optimal reactive power configuration scheme, the reactive source configuration type and the constraints to determine whether the requirements of no-difference compensation are met; if the requirements of no-difference compensation are not met, identify the nodes and their reactive sources that do not meet the requirements of no-difference compensation; adjust the parameters of the nodes and their reactive sources that do not meet the requirements of no-difference compensation and re-perform multi-objective optimization solutions; if the requirements of no-difference compensation are met, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation.

[0041] The present application also provides a storage medium, comprising: a computer executable program stored therein, wherein the computer executable program is used to be called by a processor to execute the steps of the above-mentioned multi-objective reactive power compensation planning method.

[0042] The beneficial effects of this application are:

[0043] The present application provides a multi-objective reactive power compensation planning method, including: applying K-means algorithm to cluster renewable energy output data into multiple typical scenarios; calculating the L index and L index sensitivity of each node in each typical scenario; selecting the top 10 candidate nodes with the initial value of the L index sensitivity in each typical scenario; selecting the top 5 candidate nodes with the highest frequency of occurrence among all the candidate nodes as reactive power compensation nodes; setting an objective function according to the L index improvement amount of the top 10 nodes in each scenario and the minimization of the system network loss, and using multi-objective optimization to solve the problem, and obtaining the L index and L index sensitivity of each node in each typical scenario. The optimal reactive power configuration scheme; determine the reactive source configuration type of the reactive compensation node, and the system constraints; perform simulated compensation for each typical scenario according to the optimal reactive power configuration scheme, reactive source configuration type and the constraints, and determine whether the requirements of no-difference compensation are met; if the requirements of no-difference compensation are not met, identify the nodes and their reactive sources that do not meet the requirements of no-difference compensation; adjust the parameters of the nodes and their reactive sources that do not meet the requirements of no-difference compensation and re-perform multi-objective optimization solutions; if the requirements of no-difference compensation are met, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation. This application improves the rationality of the optimization and supplementary scheme by verifying and feeding back the optimal configuration scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic diagram of the multi-objective reactive power compensation planning process in this application;

[0045] Figure 2 It is a schematic diagram of the multi-objective reactive power compensation planning model in this application;

[0046] Figure 3 It is a schematic diagram of a multi-objective reactive power compensation planning device in this application;

[0047] Figure 4 It is a schematic diagram of the multi-objective reactive power compensation planning equipment in this application. DETAILED DESCRIPTION

[0048] The present application is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present application and implement it.

[0049] Please refer to Figure 1 As shown, the present application provides a multi-objective reactive power compensation planning method, the steps of which include:

[0050] S101. Use K-means algorithm to cluster the renewable energy output data into multiple typical scenarios.

[0051] First, it is necessary to collect the renewable energy power generation data of a certain region throughout the year (8760 hours), including key indicators such as hourly power generation and power output.

[0052] The collected data is cleaned and preprocessed to remove outliers and missing values ​​to ensure the quality and integrity of the data. At the same time, the data needs to be normalized to better compare the output at different time points during the clustering process.

[0053] In the K-means clustering algorithm, the K value represents the number of clusters, that is, the number of typical scenarios to be formed in the end. In the embodiment of the present application, 96 is preferably selected as the K value, which means that the new energy output data for the whole year is clustered into 96 typical scenarios. This choice is usually based on actual needs, computing resources, and understanding of the characteristics of new energy output. Therefore, those skilled in the art can choose other values ​​as the K value according to actual needs.

[0054] At the beginning of the algorithm, K data points are randomly selected as the initial cluster centers. These centers will serve as the starting points for subsequent iterations.

[0055] For each point in the data set, calculate its distance to each cluster center (such as Euclidean distance) and assign it to the class where the closest cluster center is located. Update the cluster center of each class to the mean of all points in that class. Repeat the above steps until the cluster center no longer changes significantly or reaches the preset number of iterations, at which time the clustering is considered complete.

[0056] After clustering, 96 cluster centers and their corresponding class members (i.e., time series data of renewable energy output) are obtained. These cluster centers represent 96 typical renewable energy output scenarios.

[0057] Each typical scenario represents a typical mode or state of renewable energy output in a certain region. These scenarios may reflect the impact of different seasons, weather conditions or load demands on renewable energy output.

[0058] By analyzing the typical scenarios of renewable energy output, the complexity of subsequent analysis can be reduced. In the subsequent reactive power compensation planning, only these 96 typical scenarios need to be optimized and simulated, without considering every specific moment of the 8,760 hours throughout the year, which significantly improves planning efficiency and accuracy.

[0059] S102: Calculate the L index and L index sensitivity of each node in each typical scenario.

[0060] In the power system, the L index (or voltage stability margin index) is an important parameter to measure the static voltage stability of the system. It reflects the system's ability to maintain voltage stability under specific operating conditions when faced with load growth or reactive power changes.

[0061] L index sensitivity (LQSIj) further describes the extent to which the L index changes when the reactive power of a node (or a group of nodes) in the system changes slightly (ΔQ). This sensitivity analysis is crucial for reactive power compensation planning because it can reveal which nodes' reactive power compensation is most effective in improving system voltage stability.

[0062] For each typical scenario, it is necessary to obtain the operating status data of the system in the scenario, including the voltage, active power, reactive power, etc. of each node. These data are the basis for calculating the L index and its sensitivity.

[0063] In each scenario, the L index value of the current system state is calculated (denoted as L). The calculation of the L index usually involves complex processes such as system power flow analysis, calculation of the Jacobian matrix, and evaluation of the voltage stability margin.

[0064] In order to calculate the sensitivity of the L index, it is necessary to simulate a small change in reactive power at a node (or a group of nodes) in the power grid system. This can be achieved by changing the reactive injection amount of the node, and the change is recorded as ΔQ. Multiple simulations with different ΔQ values ​​can be performed to obtain more accurate results. After simulating the reactive power change, the L index value of the system is recalculated (recorded as L(ΔQ)). This step also requires related calculations such as power flow analysis.

[0065] Finally, the L index values ​​obtained in the above two steps are used to calculate the L index sensitivity LQSIj. The calculation formula is:

[0066] LQSIj=(LL(ΔQ)) / ΔQ

[0067] Among them, j represents the node number where the reactive power changes. This formula expresses the ratio of the change in L index to ΔQ when the reactive power of node j changes by ΔQ, that is, the sensitivity of L index to the reactive power change of the node.

[0068] Repeating the above steps for each of the 96 scenarios, the L index sensitivity of each node in each scenario can be obtained. By analyzing these sensitivity values, it is possible to identify which nodes are most critical to improving system voltage stability during reactive power compensation, thus providing an important basis for subsequent reactive power compensation planning.

[0069] S103: Select the top 10 candidate nodes with the initial sensitivity value of the L index in each typical scenario.

[0070] In reactive power compensation planning of power systems, when analyzing multiple typical scenarios of renewable energy output, a key step is to identify the nodes that are most sensitive to or have the greatest impact on reactive power compensation for each scenario.

[0071] In the above steps, for each typical scenario obtained by K-means clustering, the sensitivity calculation of the L indicator has been completed.

[0072] For each typical scenario, the corresponding node L index sensitivity list is sorted. The sorting is based on the sensitivity value, from large to small. In this way, the nodes with the highest sensitivity value will be at the front of the list, indicating that these nodes are most sensitive to reactive power compensation.

[0073] In the sorted list, the first 10 nodes are selected as the most critical nodes for reactive power compensation in this scenario. These nodes have a large initial value of L index sensitivity, which means that in the reactive power compensation process, by adjusting the reactive power of these nodes, the voltage stability of the system can be improved more effectively or the network loss of the system can be reduced.

[0074] Repeat the above steps for each scenario, and finally get a list of 10 key nodes corresponding to each of the 96 scenarios. These node lists are an important basis for subsequent reactive power compensation planning, and can be used to guide the installation location, capacity configuration, and control strategy of reactive power compensation equipment.

[0075] S104: Select the top five candidate nodes with the highest appearance frequency among all the candidate nodes as reactive power compensation nodes.

[0076] After performing L index sensitivity analysis on 96 typical scenarios, 10 nodes with larger initial values ​​have been selected for each scenario as candidate nodes that are more sensitive to reactive power compensation in that scenario. In this way, a total of 960 nodes (96 scenarios * 10 nodes / scenario) were selected. Next, in order to determine the nodes that most need reactive power compensation in the entire system, it is necessary to perform statistical analysis on these nodes to select the 5 nodes with the highest frequency as the final reactive power compensation nodes.

[0077] First, create a node frequency statistics table. This table will list all the 960 selected nodes and set a counter for each node. Then, traverse these 960 nodes, and each time a node is selected as a candidate node in a certain scene, increase its corresponding counter by 1. In this way, the counter of each node will eventually record the number of times the node appears as a candidate node in all scenes, that is, its frequency.

[0078] After completing the frequency statistics, these nodes are sorted according to their frequency of occurrence. The sorting is based on the size of the node frequency, from large to small. In this way, the node with the highest frequency will be at the front of the list.

[0079] In the sorted list, the first five nodes are selected as the final reactive power compensation nodes. These nodes appear most frequently as candidate nodes in all scenarios, which means that they have the most significant impact on the voltage stability or reactive power balance of the entire system. Therefore, reactive power compensation at these nodes will maximize the overall performance of the system.

[0080] S105. Setting an objective function according to the L index improvement of the top 10 nodes in each scenario and minimizing the system network loss, and using multi-objective optimization to obtain the optimal reactive power configuration solution for each of the typical scenarios.

[0081] In the reactive power optimization problem of the power system, it is an important step to calculate the initial value of the L index for multiple typical scenarios and determine the priority nodes for reactive power compensation in each scenario based on these calculation results. Next, using these nodes and their L indexes, combined with the system network loss as the objective function, a multi-objective optimization algorithm is used to solve the optimal reactive power configuration scheme in each typical scenario.

[0082] First, for each typical scenario, the initial value of the L indicator is calculated.

[0083] The choice of L index depends on the specific power system analysis requirements, including an index that reflects the system voltage stability (such as L-index) or an index related to reactive power reserve capacity. This step involves the calculation of system flow and the evaluation of L index based on the current state of the system.

[0084] After the initial value of L index is calculated for each scenario, the next step is to select 10 nodes with larger L index in each scenario. These nodes are considered to be the most sensitive or most influential nodes for reactive power compensation in this scenario, because their higher L index values ​​indicate that their state changes have a greater impact on the overall performance of the system.

[0085] For each scenario, two objective functions are defined: one is based on the L index of the selected 10 nodes, aiming to reduce the L index value of these nodes through reactive power compensation, thereby improving the voltage stability or reactive power reserve capacity of the system; the other is the system network loss, that is, the energy loss caused by resistance in the system. Reactive power optimization is often committed to reducing the system network loss to improve economy.

[0086] Since there are multiple objective functions (L index and system network loss), this is a typical multi-objective optimization problem. Multi-objective optimization algorithms (such as NSGA-II, MOEA / D, etc.) are used to solve this type of problem. These algorithms can consider multiple objectives simultaneously in one run and generate a set of non-dominated solutions (Pareto optimal solution set).

[0087] In the multi-objective optimization process, the algorithm attempts to adjust the configuration of the reactive power compensation equipment (such as capacity, location, etc.) to find the best balance between reducing the L index and system network loss.

[0088] Finally, the algorithm will output a set of Pareto optimal solutions, each of which represents a possible optimal reactive power configuration scheme. These schemes achieve different balances between reducing the L index and reducing system network losses.

[0089] S106: Determine the reactive power source configuration type of the reactive power compensation node and system constraints.

[0090] After determining the reactive power source configuration of each node in the power system, in order to maximize the effect of reactive power compensation and ensure the stable operation of the system, it is necessary to consider the output characteristics of each reactive power source, so as to achieve the coordination between reactive power sources. This process aims to achieve the best reactive power compensation solution by optimizing the working status and output distribution of different reactive power sources (such as generators, static VAR compensators SVC, static synchronous compensators STATCOM, etc.). 1. Reactive power source characteristics analysis

[0091] Conduct detailed characteristic analysis of reactive power sources at each node. This includes understanding the regulation range, response speed, control accuracy, economic cost, and possible technical limitations of each reactive power source. For example, generators usually have a large reactive power regulation capability, but their regulation speed may be limited by the stability of the unit and the performance of the excitation system; while power electronic devices such as SVC and STATCOM have faster response speeds and higher control accuracy, but may be limited by cost and maintenance requirements.

[0092] Based on the system operation status and the prediction of future load changes, the reactive power demand of each node in different time periods is evaluated. Then, according to the output characteristics of the reactive power source and the system constraints (such as voltage stability, reactive power balance, transmission line capacity, etc.), the reactive power demand is reasonably allocated to each reactive power source. This step requires comprehensive consideration of the overall performance and economy of the system to ensure the efficiency and effect of reactive power compensation.

[0093] In order to achieve the coordination of reactive power sources, a set of coordinated control strategies needs to be developed. This strategy should be able to dynamically adjust the output distribution of each reactive power source according to the changes in the system operating status to ensure that the reactive power demand of the system is met and the voltage stability is maintained. The coordinated control strategy may include optimization algorithms based on centralized control (such as linear programming, nonlinear programming, genetic algorithms, etc.), or may include intelligent control methods based on decentralized control (such as fuzzy control, neural network control, etc.).

[0094] S107: Perform simulated compensation for each of the typical scenarios according to the optimal reactive power configuration scheme, reactive power source configuration type and the constraint conditions to determine whether the requirement of zero-difference compensation is met.

[0095] For the selected compensation nodes in the power system, it is necessary to clarify the type of reactive source configured on each node. Reactive source types include capacitors (Capacitor Banks), synchronous condensers (Synchronous Condenser s), static var compensators (SVC) and static synchronous compensators (SVG). These reactive sources have their own characteristics, such as capacitors have low cost but limited adjustment range, condensers and SVG have faster response speed and wider adjustment range, and SVC is between them.

[0096] Next, we need to analyze the constraints of each reactive power source. These constraints may include but are not limited to:

[0097] Reactive power regulation range: Each reactive power source has its own output reactive power range.

[0098] Response speed: Different reactive power sources have different response speeds to system changes.

[0099] Economic cost: including initial investment, operation and maintenance costs, etc.

[0100] Technical limitations: such as maximum current, voltage limitations, etc.

[0101] After identifying the reactive power source type and constraints, the above multi-objective optimization algorithms (such as NSGA-II, MOEA / D, etc.) are used to solve the optimal reactive power configuration scheme for each typical scenario. This step aims to find the optimal reactive power configuration capacity that meets multiple objectives such as system voltage stability, reactive power balance, and economy.

[0102] After obtaining the optimal reactive power configuration capacity in each typical scenario, it is necessary to combine these capacities with the reactive power source type and its characteristics of the corresponding node. Specifically, it is to determine the input quantity and output distribution of each reactive power source based on the reactive power demand of each node (i.e. the optimal configuration capacity) and the available reactive power source type.

[0103] Finally, based on the above analysis, differential compensation is implemented. Differential compensation means that while meeting the reactive power demand of the system, the error or imbalance caused by reactive power compensation is minimized. This usually requires precise control of the output of each reactive power source to ensure that they can work together in a predetermined ratio and sequence to achieve the best compensation effect.

[0104] S108: If the requirement of the zero-difference compensation is not met, identify the nodes and their reactive power sources that do not meet the requirement of the zero-difference compensation.

[0105] When performing reactive power compensation in a power system, it is crucial to ensure that each node can achieve a state of zero-difference compensation. However, in actual operation, it may be encountered that the reactive power source configured at some nodes cannot meet the compensation requirements, such as insufficient capacity of the reactive power source or failure to reach its compensation lower limit.

[0106] After completing the preliminary setting of the reactive power compensation scheme, the reactive power compensation effect of each node is verified. This includes checking whether the reactive power source (such as capacitors, phase regulators, SVCs, SVGs, etc.) configured at the node can meet the reactive power demand in the current scenario. The basis for the evaluation is the reactive load of the node, the voltage stability requirements, and the output characteristics and constraints of the reactive power source.

[0107] During the verification process, if it is found that the reactive power compensation equipment of a node cannot meet the compensation requirements, that is, there is insufficient capacity or the compensation lower limit cannot be reached, then the node is considered as a node that does not meet the requirements. This may be because the reactive load of the node is large, or the configured reactive power source type and capacity do not match.

[0108] S109, adjusting the parameters of the nodes and their reactive sources that do not meet the zero-difference compensation requirements and re-performing the multi-objective optimization solution.

[0109] For nodes that do not meet the requirements, further constraint checks are performed. This includes checking whether the reactive power source of the node is subject to any technical or economic restrictions, such as maximum current restrictions, voltage stability restrictions, cost budget restrictions, etc. These restrictions will affect the output capacity and configuration of the reactive power source.

[0110] After confirming that the node does not meet the requirements and is subject to constraints, the reactive power compensation solution is re-solved according to the current scenario. This step usually involves the following considerations:

[0111] Adjust the reactive power source type: If the currently configured reactive power source type is not suitable for the node's needs, you can consider replacing it with a more suitable reactive power source type.

[0112] Increase reactive power source capacity: If insufficient capacity is the main reason, consider increasing the capacity of reactive power sources at that node or nearby nodes.

[0113] Optimize reactive power source layout: By optimizing the layout of reactive power sources, the efficiency and effect of reactive power compensation can be improved. For example, configuring reactive power sources closer to the load center can reduce the transmission loss of reactive power in the power grid.

[0114] Consider multi-scenario optimization: Since the operating conditions of the power system are complex and changeable, the reactive power demand under multiple typical scenarios is considered, and the reactive power compensation scheme is comprehensively optimized.

[0115] If the requirement of zero-difference compensation is met in S110, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation.

[0116] When reactive power compensation is performed in a power system, if the error compensation can be achieved within the specified range, the resulting compensation scheme is regarded as a compensation scheme for the coordinated operation of multiple types of reactive sources at each node.

[0117] Evaluate whether the current reactive power compensation scheme can achieve error compensation within the specified range. This usually involves analysis and comparison of the actual system operation data to determine the reactive power supply and demand balance, voltage fluctuations and possible error range under the current compensation scheme. If the error is within an acceptable range, the current compensation scheme can be considered effective.

[0118] According to the above technical solution, the present application designs a multi-objective reactive power compensation planning model.

[0119] Please refer to Figure 2As shown, the model includes an upper layer model and a lower layer model, the upper layer model is used to process steps S101 to S105 and S109, and the lower layer model is used to process steps S106 to S108 and step S110.

[0120] When the lower model reaches step S109, it jumps to the upper model and optimizes the solutions for the nodes that do not meet the zero-error compensation requirements according to the new node parameters.

[0121] The interaction between the upper layer model and the lower layer model completes the execution of the above steps S101 to S110.

[0122] Please refer to Figure 3 As shown, the present application also provides a multi-objective reactive power compensation planning device, comprising:

[0123] The scenario module 201 uses the K-means algorithm to cluster the renewable energy output data into multiple typical scenarios;

[0124] A calculation module 202 calculates the L index and L index sensitivity of each node in each typical scenario;

[0125] The selection module 203 selects the top 10 candidate nodes with the initial sensitivity value of the L index in each typical scenario;

[0126] The node module 204 selects the top five candidate nodes with the highest appearance frequency among all the candidate nodes as reactive power compensation nodes;

[0127] The optimization module 205 sets the objective function according to the L index improvement of the top 10 nodes of the L index in each scenario and the minimization of the system network loss, and uses multi-objective optimization to obtain the optimal reactive power configuration solution under each typical scenario;

[0128] The condition module 206 determines the reactive source configuration type of the reactive compensation node and the system constraint conditions;

[0129] The verification module 207 performs simulated compensation for each of the typical scenarios according to the optimal reactive power configuration scheme, reactive power source configuration type and the constraint conditions to determine whether the requirement of zero-difference compensation is met;

[0130] Identification module 208, if the requirement of zero-error compensation is not met, identifies the node and its reactive power source that do not meet the requirement of zero-error compensation;

[0131] Feedback module 209 adjusts the parameters of nodes and their reactive sources that do not meet the requirements of zero-difference compensation and performs multi-objective optimization again;

[0132] The execution module 210 applies the reactive power compensation scheme to the actual power system to perform reactive power compensation if the requirement of zero-difference compensation is met.

[0133] Please refer to Figure 4 As shown, the present application also provides a multi-objective reactive power compensation planning device, including:

[0134] Memory 301;

[0135] Processor 302, used to retrieve the computer executable program of the above-mentioned multi-objective reactive power compensation planning method from the memory, and execute: applying K-means algorithm to cluster the renewable energy output data into multiple typical scenarios; calculating the L index and L index sensitivity of each node in each typical scenario; selecting the top 10 candidate nodes with the initial value of the L index sensitivity in each typical scenario; selecting the top 5 candidate nodes with the highest frequency of occurrence among all the candidate nodes as reactive power compensation nodes; setting the objective function according to the L index improvement amount of the top 10 nodes in each scenario and the minimization of the system network loss, and using multi-objective optimization to solve the problem; The optimal reactive power configuration scheme for each of the typical scenarios is obtained; the reactive source configuration type of the reactive compensation node and the system constraints are determined; according to the optimal reactive power configuration scheme, the reactive source configuration type and the constraints, simulated compensation is performed for each of the typical scenarios to determine whether the requirements for no-difference compensation are met; if the requirements for no-difference compensation are not met, the nodes and their reactive sources that do not meet the requirements for no-difference compensation are identified; the parameters of the nodes and their reactive sources that do not meet the requirements for no-difference compensation are adjusted and the multi-objective optimization solution is re-performed; if the requirements for no-difference compensation are met, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation.

[0136] The present application also provides a storage medium, comprising: a computer executable program stored therein, wherein the computer executable program is used to be called by a processor to execute the steps of the above-mentioned multi-objective reactive power compensation planning method.

[0137] The embodiments and / or implementation methods described above are only used to illustrate the preferred embodiments and / or implementation methods for realizing the technology of the present invention, and are not intended to limit the implementation methods of the technology of the present invention in any form. Any technical personnel in this field may make slight changes or modifications to other equivalent embodiments without departing from the scope of the technical means disclosed in the content of the present invention, but they should still be regarded as technologies or embodiments that are essentially the same as the present invention.

[0138] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and its core ideas of this application. The above is only the preferred implementation method of this application. It should be pointed out that due to the limitations of textual expression and the objective existence of infinite specific structures, ordinary technicians in this technical field can make some improvements, modifications or changes without departing from the principles of this application, and combine the above technical features in an appropriate manner; the improvements, modifications, changes or combinations, or the direct application of the concept and technical solution of the invention to other occasions without improvement, should be regarded as the scope of protection of this application.

Claims

1. A multi-objective reactive power compensation planning method, characterized in that: include: The K-means algorithm is used to cluster the renewable energy output data into multiple typical scenarios; Calculate the L index and L index sensitivity of each node in each typical scenario; Select the top 10 candidate nodes with the initial value of L index sensitivity in each typical scenario; The top five candidate nodes with the highest appearance frequency among all the candidate nodes are selected as reactive power compensation nodes; The objective function is set according to the L index improvement of the top 10 nodes in each scenario and the minimization of the system network loss, and a multi-objective optimization solution is used to obtain the optimal reactive power configuration solution for each typical scenario; Determining the reactive source configuration type of the reactive compensation node and system constraints; According to the optimal reactive power configuration scheme, reactive power source configuration type and the system constraints, simulate compensation for each typical scenario to determine whether the zero-difference compensation requirement is met; If the no-difference compensation requirement is not met, the nodes and their reactive power sources that do not meet the no-difference compensation requirement are identified; Adjust the parameters of nodes and their reactive sources that do not meet the requirements of zero-difference compensation and perform multi-objective optimization again; If the no-difference compensation requirements are met, the reactive power compensation scheme is applied to the actual power system for reactive power compensation.

2. The multi-objective reactive power compensation planning method according to claim 1, characterized in that: The L index sensitivity LQSIj is calculated as follows: LQSIj = (LL (ΔQ)) / ΔQ; Where L is the L index value before reactive power optimization, L(ΔQ) is the L index value after reactive power compensation, and ΔQ is the reactive power compensation amount.

3. The multi-objective reactive power compensation planning method according to claim 1, characterized in that: The reactive source configuration types include one or more of capacitors, phase regulators, SVCs, and SVGs.

4. The multi-objective reactive power compensation planning method according to claim 1, characterized in that: If the requirements of no-difference compensation are met, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation, including: Monitor the static voltage stability of the power system, evaluate the reactive power compensation effect, and adjust the optimization strategy based on the evaluation results.

5. The multi-objective reactive power compensation planning method according to claim 1, characterized in that: The K-means algorithm is used to cluster the renewable energy output data into several typical scenarios, including: According to the time series characteristics of renewable energy output, load changes and weather factors, the K-means algorithm is used to cluster the renewable energy output data.

6. The multi-objective reactive power compensation planning method according to claim 1, characterized in that: After selecting the top 10 candidate nodes of the initial value of L index sensitivity in each typical scenario, including: A voltage stability assessment is performed on the candidate nodes to ensure that the candidate nodes can improve the overall voltage stability of the system after reactive power compensation.

7. The multi-objective reactive power compensation planning method according to claim 1, characterized in that: Also includes: A system model is established in the power system simulation software, including the renewable energy power generation model, load model, and transmission network model, to more accurately simulate the operation of the actual power system.

8. A multi-objective reactive power compensation planning device, characterized in that: include: The scenario module uses the K-means algorithm to cluster the renewable energy output data into multiple typical scenarios; A calculation module, calculating the L index and L index sensitivity of each node in each typical scenario; A selection module selects the top 10 candidate nodes with the initial sensitivity value of the L indicator in each typical scenario; A node module, taking the top 5 candidate nodes with the highest appearance frequency among all the candidate nodes as reactive power compensation nodes; The optimization module sets the objective function according to the L index improvement of the top 10 nodes in each scenario and the minimization of the system network loss, and uses multi-objective optimization to obtain the optimal reactive power configuration solution under each typical scenario; A condition module, determining the reactive source configuration type of the reactive compensation node and system constraints; A verification module, which performs simulated compensation for each of the typical scenarios according to the optimal reactive power configuration scheme, reactive power source configuration type and the constraint conditions, and determines whether the requirement of zero-difference compensation is met; The identification module identifies the nodes and their reactive power sources that do not meet the requirements of the differential compensation if the requirements of the differential compensation are not met; Feedback module, adjusts the parameters of nodes and their reactive sources that do not meet the requirements of zero-difference compensation and performs multi-objective optimization again; The execution module, if the requirements of no-difference compensation are met, the reactive power compensation scheme is applied to the actual power system to perform reactive power compensation.

9. A multi-objective reactive power compensation planning device, characterized in that: include: Memory; A processor, used to retrieve from the memory a computer executable program of a multi-objective reactive power compensation planning method as described in any one of claims 1 to 7, and execute: clustering the renewable energy output data using a K-means algorithm to cluster into a plurality of typical scenarios; calculating the L index and L index sensitivity of each node in each of the typical scenarios; selecting the top 10 candidate nodes with the initial value of the L index sensitivity in each of the typical scenarios; The top 5 candidate nodes with the highest frequency among all the candidate nodes are selected as reactive compensation nodes; the objective function is set according to the L index improvement of the top 10 nodes in each scenario and the minimization of the system network loss, and a multi-objective optimization solution is used to obtain the optimal reactive configuration scheme for each typical scenario; the reactive source configuration type of the reactive compensation node and the system constraints are determined; according to the reactive optimal configuration scheme, the reactive source configuration type and the constraints, simulated compensation is performed for each typical scenario to determine whether the requirements for no-difference compensation are met; if the requirements for no-difference compensation are not met, the nodes and their reactive sources that do not meet the requirements for no-difference compensation are identified; the parameters of the nodes and their reactive sources that do not meet the requirements for no-difference compensation are adjusted and the multi-objective optimization solution is performed again; If the requirements of zero-difference compensation are met, the reactive power compensation scheme is applied to the actual power system for reactive power compensation.

10. A storage medium, characterized in that: include: A computer executable program is stored, and the computer executable program is used to be called by a processor to execute the steps of a multi-objective reactive power compensation planning method as described in any one of claims 1 to 7.

Images