Method and system for establishing comprehensive load model containing distributed power supply

Abstract

The invention discloses a method and a system for establishing a comprehensive load model containing a distributed power supply, and relates to the technical field of load model establishment, after the establishment system obtains power system information through an API interface, a topological graph model of a power system is established, each component in the power system is abstracted as a node and an edge in the topological graph model, and the topological graph model is established; after simulation software is used for conducting simulation detection on the topological graph model, a target value of the topological graph model is calculated through an objective function, an optimal path between nodes is updated through a path algorithm, current distribution of edges is adjusted through a graph cut algorithm, power loss is reduced, overload is avoided, the steps are repeated for multiple iterations, and when convergence conditions are met, the optimal path is obtained. And outputting the final topological graph model as a comprehensive load model. The system can optimize the targets at the same time through a graph optimization algorithm, improves the overall benefits of the system, and overcomes the defect that the complexity of the power system cannot be comprehensively considered in single target optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of load model establishment, and in particular to a method and system for establishing a comprehensive load model including distributed power sources. Background Art

[0002] Load models are mainly used in modern power systems, which include traditional loads (such as residential, commercial, and industrial electricity) and distributed power sources (such as solar energy, wind energy, small hydropower, etc.). With the popularization of distributed power sources, especially the access of renewable energy, the load characteristics of the power system have changed. The load model can better simulate and predict the operating status of the system, thereby improving the scheduling, optimization, and safety of the power system.

[0003] The prior art has the following defects: Traditional power load forecasting and generation scheduling methods only focus on a single goal (such as load demand forecasting or generation resource scheduling), without considering the balance and interaction between multiple goals. For example, the load forecasting accuracy often cannot match the generation scheduling requirements, especially in power grids with a high proportion of distributed power sources. The volatility of distributed power sources makes load scheduling and generation arrangements more complicated, reducing the efficiency and stability of power system operation. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for establishing a comprehensive load model containing distributed power sources. Through a graph optimization algorithm, these objectives can be optimized simultaneously to improve the overall efficiency of the system, solve the defect that single-objective optimization cannot fully consider the complexity of the power system, and make the operation of the power system more balanced and efficient.

[0005] In order to achieve the above object, the present invention provides the following technical solution: a method for establishing a comprehensive load model containing distributed power sources, the establishment method comprising the following steps: S1: After the system is established to obtain power system information through the API interface, a topological graph model of the power system is established, and each component in the power system is abstracted into nodes and edges in the topological graph model; S2: After using simulation software to simulate and test the topology model, the target value of the topology model is calculated using the objective function, the optimal path between nodes is updated through the path algorithm, and the current distribution of the edge is adjusted through the graph cut algorithm; S3: Repeat step S2 for multiple iterations, and when the convergence condition is met, output the final topology model as the comprehensive load model.

[0006] In a preferred embodiment, the target value of the topology graph model is calculated using an objective function, and the objective function expression is: , where is the target value, is the power system stability index, is the load deviation amplitude, is the power loss, , , are the power system stability index, load deviation amplitude and power loss adjustment coefficient respectively, and , , Both are greater than 0.

[0007] In a preferred embodiment, step S2 is repeated for multiple iterations to determine whether the convergence condition is met, including the following steps: Repeat the target value calculation, optimal path update and current distribution adjustment steps for the current topology model for multiple iterations. When the number of iterations is equal to the number threshold or the target value of the topology model after a certain update is greater than or equal to the performance threshold, it is determined that the convergence condition is met.

[0008] In a preferred embodiment, when the convergence condition is met, the final topology model is output as a comprehensive load model, including the following steps: When the convergence condition is met, the topology model obtained in each iteration is output, and the topology models obtained in each iteration are sorted from large to small according to the target value, and a model sorting table is obtained. The topology model ranked first in the model sorting table is selected as the final topology model, and the final topology model is output as the comprehensive load model; The comprehensive load model is applied in the distribution network. After recording the target value of the comprehensive load model at multiple time points, it is analyzed whether the comprehensive load model needs to be replaced. If it needs to be replaced, the corresponding comprehensive load model is selected according to the model sorting table.

[0009] In a preferred embodiment, the power loss is calculated as follows: , where For Node and nodes The current value between For Node and nodes The resistance value of the line between them; The calculation logic of the power system stability index is: real-time monitoring of the output current fluctuation value of the distributed power supply node, when the output current fluctuation value is greater than the fluctuation threshold, the power system stability index is calculated, and the expression is: , where is the power system stability index, , Node and nodes The voltage value, , For Node The admittance matrix elements between , Node and nodes The phase angle of .

[0010] In a preferred embodiment, the calculation logic of the load deviation amplitude is: obtain the actual load of each load node in the current topology model, and calculate the load deviation of each load node, the expression is: , where is the load deviation, is the actual load, is the standard load. After obtaining the load deviation of each load node, the mean value of the load deviation and the standard deviation of the load deviation are calculated. The expression is: , where is the load deviation standard deviation, is the mean value of load deviation, For the The load deviation of each load node, is the number of load nodes, and the load deviation amplitude is obtained by dividing the load deviation mean by the load deviation standard deviation.

[0011] In a preferred embodiment, after the system acquires the power system information through the API interface, a topological graph model of the power system is established, and each component in the power system is abstracted into nodes and edges in the topological graph model, including the following steps: Establish a system to obtain various information of the power system in real time through the API interface, including demand data of load points, power generation data of distributed power sources, data of power transmission lines, and operating status of the power system; Each node represents a load point in the power system, including residential load points, industrial load points, commercial load points, and factory load points. The attributes of each load point include load value, historical data, distributed power type, power generation, power generation historical data, and power generation mode. The edges represent the power transmission paths connecting different nodes, connecting the load points and the power supply points, and each edge represents the power flow relationship between the nodes; The topological graph model of the power system consists of nodes and edges, forming a directed graph, where each edge represents the direction and restriction of power flow.

[0012] In a preferred embodiment, the integrated load model is applied to the distribution network, and after the target value of the integrated load model is recorded at multiple time points, analyzing whether the integrated load model needs to be replaced includes the following steps: After obtaining the target values ​​of the comprehensive load model at multiple time points, the maximum target value is selected as the target peak value, the target deviation is calculated based on the performance threshold and the target peak value, the target values ​​at multiple time points are integrated to obtain the target cumulative index, and the target values ​​at multiple time points are analyzed to obtain the target fluctuation amplitude. After the target deviation, target cumulative index and target fluctuation amplitude are amplified, the selection coefficient of the comprehensive load model is obtained by comprehensive calculation, and the expression is: , where is the selection coefficient, is the target deviation, is the target cumulative index, is the target fluctuation amplitude, , , are the information gains of target deviation, target cumulative index and target fluctuation amplitude respectively, and , , All are greater than 0; The obtained selection coefficient is compared with the preset selection threshold. The selection threshold is used to determine whether the comprehensive load model needs to be replaced. If the selection coefficient is greater than or equal to the selection threshold, the overall performance of the comprehensive load model applied in the distribution network is analyzed to be good, and it is determined that the comprehensive load model does not need to be replaced. If the selection coefficient is less than the selection threshold, the overall performance of the comprehensive load model applied in the distribution network is analyzed to be poor, and it is determined that the comprehensive load model needs to be replaced.

[0013] In a preferred embodiment, the target deviation is calculated based on the performance threshold and the target peak value, and the calculation logic is: the target deviation is obtained by subtracting the target peak value from the performance threshold value; The target cumulative index is obtained by integrating the target values ​​at multiple time points. The expression is: , where represents the operation time of the comprehensive load model in the distribution network, represents the target value of the comprehensive load model at the time instant; After analyzing the target values ​​at multiple time points, the target fluctuation amplitude is obtained. The expression is: , where is the number of time points, For the The target value at a time point, is the target mean.

[0014] A comprehensive load model building system including distributed power sources, including a topology model building module, an optimization module, and an iterative output module; Topology model building module: After obtaining power system information through the API interface, a topology model of the power system is established, and each component in the power system is abstracted into nodes and edges in the topology model; Optimization module: After using simulation software to simulate and detect the topology model, the target value of the topology model is calculated using the objective function, the optimal path between nodes is updated through the path algorithm, and the current distribution of the edge is adjusted through the graph cut algorithm; Iterative output module: Repeat the steps in the optimization module for multiple iterations. When the convergence conditions are met, the final topology model is output as the comprehensive load model.

[0015] In the above technical solution, the technical effects and advantages provided by the present invention are: 1. The present invention establishes a topological model of the power system after acquiring power system information through an API interface by establishing a system, abstracting various components in the power system into nodes and edges in the topological model, and using simulation software to simulate and detect the topological model. Then, the target value of the topological model is calculated using an objective function, the optimal path between nodes is updated through a path algorithm, and the current distribution of edges is adjusted through a graph cut algorithm to reduce power loss and avoid overload. The above steps are repeated for multiple iterations, and when the convergence conditions are met, the final topological model is output as a comprehensive load model. The system is established by defining a comprehensive objective function and taking into account multiple optimization objectives, such as minimizing load prediction errors, reducing power losses, optimizing power flow, etc. Through the graph optimization algorithm, these objectives can be optimized simultaneously to improve the overall efficiency of the system and solve the defect that single-objective optimization cannot fully consider the complexity of the power system; 2. After obtaining the target values ​​of the comprehensive load model at multiple time points, the present invention selects the maximum target value as the target peak value, calculates the target deviation based on the performance threshold and the target peak value, integrates the target values ​​at multiple time points to obtain the target cumulative index, and obtains the target fluctuation amplitude after analyzing the target values ​​at multiple time points. After gaining the target deviation, the target cumulative index and the target fluctuation amplitude, the selection coefficient of the comprehensive load model is obtained by comprehensive calculation. By comprehensively analyzing the application status of the comprehensive load model at multiple time points, it is determined whether the comprehensive load model needs to be replaced, thereby further ensuring the stability of the distribution network operation and improving the efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0019] Example 1: Please refer to Figure 1 As shown, the method for establishing a comprehensive load model containing distributed power sources described in this embodiment includes the following steps: S1: After the system is established to obtain power system information through the API interface, a topological graph model of the power system is established, and each component in the power system is abstracted into nodes and edges in the topological graph model; S2: After using simulation software to simulate and test the topology model, the target value of the topology model is calculated using the objective function, the optimal path between nodes is updated through the path algorithm, and the current distribution of the edge is adjusted through the graph cut algorithm; S3: Repeat step S2 for multiple iterations, and when the convergence condition is met, output the final topology model as the comprehensive load model.

[0020] Specifically, after the system is established to obtain power system information through the API interface, a topological graph model of the power system is established, and each component in the power system (load, distributed power source, transmission line, etc.) is abstracted into nodes and edges in the topological graph model. Each node represents a load point or distributed power source point in the power system, and the edge represents the power transmission relationship between nodes. Node definition: Each node can be a load point (such as a home, factory, etc.) or a distributed power source (such as photovoltaic, wind power, etc.). The node can store its load value, power generation, historical data and other information. Edge definition: The edge represents the power transmission path, connecting different nodes. The weight of the edge can represent the transmission capacity of the line, current limit, etc. After the topological graph model is simulated and tested by the simulation software, the target value of the topological graph model is calculated using the objective function, the optimal path between the nodes is updated through the path algorithm, and the current distribution of the edge is adjusted through the graph cut algorithm to reduce power loss and avoid overload. Repeat the simulation detection, target value calculation and optimal path update steps for multiple iterations. When the convergence conditions are met, the final topological graph model is output as a comprehensive load model.

[0021] This application establishes a topological model of the power system after acquiring power system information through an API interface, abstracts each component in the power system into nodes and edges in the topological model, uses simulation software to simulate and test the topological model, and then uses an objective function to calculate the target value of the topological model. The optimal path between nodes is updated through a path algorithm, and the current distribution of edges is adjusted through a graph cut algorithm to reduce power loss and avoid overload. The above steps are repeated for multiple iterations. When the convergence conditions are met, the final topological model is output as a comprehensive load model. This established system takes into account multiple optimization objectives by defining a comprehensive objective function, such as minimizing load prediction errors, reducing power losses, optimizing power flows, etc. Through the graph optimization algorithm, these objectives can be optimized simultaneously to improve the overall efficiency of the system and solve the defect that single-objective optimization cannot fully consider the complexity of the power system.

[0022] With the widespread application of distributed power sources (such as photovoltaics, wind power, etc.) and the increasing complexity of power systems, traditional power dispatching methods can no longer effectively cope with the challenges brought by power load and power generation fluctuations. By establishing a topological graph model of the power system and combining the path algorithm and graph cut algorithm to optimize current distribution, the stability, reliability and efficiency of the power system can be significantly improved. This solution is particularly suitable for load dispatching and power transmission optimization in power networks with a large number of distributed power sources (such as smart grids, microgrids, renewable energy concentrated areas, etc.).

[0023] 1) Optimal dispatch in smart grid: Scenario Overview: Smart grid is a modern power system that optimizes power transmission and management through information and communication technology. In smart grid, distributed power sources (such as photovoltaics, wind power, etc.) are widely deployed, while traditional power grids need to deal with the fluctuations of these distributed power sources and daily load demands. In order to ensure the stability of power transmission, reduce power loss, and prevent line overload, the power system needs to be accurately dispatched according to the load and power generation conditions at every moment.

[0024] Application steps: Data collection and topology map construction: The system obtains real-time information about each component in the power system through the API interface, including load points, distributed power sources, transmission lines, etc.

[0025] The system establishes a topological graph model based on the collected data, taking each load point and distributed power source point as a node, and the transmission line as an edge. The weight of the edge represents the current carrying capacity, voltage limit, etc.

[0026] Simulation and objective function calculation: The constructed topology model is simulated using simulation software to simulate power flow, load changes and fluctuations of distributed power sources.

[0027] Calculate the objective function and evaluate the power transmission efficiency, power loss, line load and other indicators of the current topology model.

[0028] Path algorithm optimization: According to the simulation results, the path algorithm is used to calculate the optimal current transmission path from the power generation node to the load node.

[0029] By continuously adjusting the current transmission path, the shortest and most evenly loaded path is selected to ensure stable and efficient power flow.

[0030] Graph cut algorithm adjusts current distribution: Using the graph cut algorithm, the current distribution between nodes in the power system is optimized to ensure that the current does not exceed the carrying capacity of the line.

[0031] The graph cut algorithm can avoid overload of certain lines in the power system and reduce power loss in the system.

[0032] Multiple iterations and convergence: The topology model is updated through repeated iterations to adjust the current path and current distribution until the system reaches the optimal state and meets the convergence conditions.

[0033] Output the optimized topology model: When the optimization goal is achieved, the system outputs the final topology model as a comprehensive load model to guide the actual scheduling and operation of the power system.

[0034] Specific applications: Dispatching control: Based on real-time power demand, power generation fluctuations of distributed power sources and the carrying capacity of the line, the system can automatically adjust the direction of current flow to achieve efficient power dispatching.

[0035] Optimize power distribution: Through the graph cut algorithm, ensure that the load of each line in the power system is evenly distributed, and avoid power outages or failures caused by overloading of some lines.

[0036] Power load balancing: For power systems with a large amount of renewable energy, the power load will be affected by weather and time. The graph cut algorithm can dynamically adjust the current distribution according to the load changes to maintain system stability.

[0037] 2) Power optimization management in microgrids: Scenario Overview: A microgrid is a local power system that usually includes multiple distributed power sources and loads. It can be connected to the main grid or operate in independent mode. In a microgrid, how to effectively manage power flow, especially how to balance the instability of distributed power sources (such as wind power and photovoltaic power generation fluctuations) and power demand, is a challenge.

[0038] Application steps: Data collection and topology map construction: The information of each node in the microgrid (such as distributed power sources, loads, storage devices, substations, etc.) is obtained through the API interface to build a topological model of the microgrid.

[0039] Simulation and objective function calculation: Simulate microgrid models to calculate power flow, load distribution, battery storage status, and more.

[0040] Calculate the objective function and evaluate the system's stability, efficiency, power loss and other indicators.

[0041] Path Algorithm and Current Optimization: Use path algorithms to calculate the optimal power transmission path to ensure the efficient flow of power from distributed generation to loads.

[0042] Graph cut algorithm optimizes current distribution: The current distribution within the microgrid is adjusted through the graph cut algorithm to ensure that each transmission line is not overloaded and reduce power loss.

[0043] Multiple iterations and convergence: Perform multiple simulations and path optimizations to gradually adjust the current path and load distribution until the system optimization goal is achieved.

[0044] Output the final topology model: Finally, the optimized topology model is output for real-time scheduling and load management of the microgrid.

[0045] Specific applications: Current flow optimization: In a microgrid, distributed power sources (such as photovoltaic power generation and wind power generation) will fluctuate due to environmental influences. The graph cut algorithm can help optimize the current flow and ensure the stable transmission of electricity in the power grid.

[0046] Load management: Based on the load demand and the power generation of distributed power sources, the graph cut algorithm dynamically adjusts the power transmission path to achieve load balance.

[0047] Fault recovery: When a fault occurs in a microgrid, the system can quickly restore power supply by updating the topology map and optimizing current distribution.

[0048] 3) Example of power dispatch in areas with concentrated renewable energy sources: Scenario Overview: Certain areas (such as those rich in solar and wind energy resources) will deploy a large number of renewable energy equipment, such as photovoltaic arrays, wind farms, etc. The output of these renewable energy sources is volatile, and how to manage these fluctuations and balance the grid load is a technical challenge.

[0049] Application steps: Data collection and topology map construction: The power system information in the region is obtained through the API interface, and a topology model including multiple distributed power sources (such as photovoltaic and wind power) and load points is constructed.

[0050] Simulation and objective function calculation: Simulate the power system to simulate the flow of electricity under different time periods and weather conditions.

[0051] Calculate objective functions and evaluate power system stability, power loss, overload risk, etc.

[0052] Path Algorithm and Current Optimization: Use routing algorithms to optimize power transmission paths and ensure maximum efficiency of the power system.

[0053] Graph cut algorithm optimizes current distribution: The current distribution is adjusted through the graph cut algorithm to ensure that all lines in the system are not overloaded and reduce power loss.

[0054] Multiple iterations and convergence: The current distribution is updated after each simulation, ensuring that the system is continuously optimized until convergence criteria are met.

[0055] Output the final topology model: Output the final power system topology diagram as the basis for power dispatch and optimization.

[0056] Specific applications: Responding to power generation fluctuations: The graph cut algorithm can dynamically adjust current distribution to ensure stable operation of the power system when photovoltaic and wind power generation fluctuates.

[0057] Reduce power loss: Reduce power loss in the power system and improve energy efficiency by optimizing paths and currents.

[0058] Avoid line overload: The graph cut algorithm is used to adjust the current distribution to prevent some lines from being overloaded due to carrying too much current.

[0059] Summarize: This technical solution can achieve efficient power transmission and optimization in different power system application scenarios, especially in smart grids, microgrids and areas with concentrated renewable energy sources with a large number of distributed power sources. This solution can effectively solve the power loss and line overload problems in the power system, optimize power dispatching, and improve the stability and efficiency of the system.

[0060] Example 2: After the system obtains the power system information through the API interface, a topological graph model of the power system is established, and each component in the power system (load, distributed power source, transmission line, etc.) is abstracted as nodes and edges in the topological graph model. Each node represents a load point or distributed power source point in the power system, and the edge represents the power transmission relationship between nodes. Node definition: Each node can be a load point (such as a home, factory, etc.) or a distributed power source (such as photovoltaic, wind power, etc.). The node can store its load value, power generation, historical data and other information. Edge definition: The edge represents the power transmission path, connecting different nodes. The weight of the edge can represent the transmission capacity of the line, current limit, etc., including the following steps: API interface to obtain power system information: Interface definition: The system obtains various information of the power system in real time through the API interface, including but not limited to: Demand data of load points (such as household and factory electricity load) Power generation data of distributed power sources (such as photovoltaic, wind power, etc.) Data on power transmission lines (such as current limits, power losses, line capacity, etc.) The operating status of the system (such as grid topology changes, equipment failures, real-time power flow, etc.) Data update: Get updated power system data regularly or dynamically based on demand to ensure the timeliness of the system.

[0061] Power system topology modeling: Node definition: Load node: Each node represents a load point in the power system, such as residential areas (household loads), industrial areas, commercial areas, factories, etc. The attributes of each load node include: Load value: current load demand (such as power, energy, etc.).

[0062] Historical data: load change history, forecast data, demand patterns, etc.

[0063] Distributed power generation node: Each node represents a distributed power generation in the power system, such as solar photovoltaic power generation, wind power generation, energy storage device, etc. The attributes of each distributed power generation node include: Power generation: the current power generation capacity of distributed power sources.

[0064] Power Generation History: Historical data on power generation.

[0065] Power generation mode: the impact of factors such as weather changes and seasonal fluctuations on power generation capacity.

[0066] Edge definition: Power transmission edge: The edge represents the power transmission path connecting different nodes, connecting the load point and the power source point. Each edge represents the power flow relationship between nodes.

[0067] Edge weight: The edge weight can contain multiple information: Current capacity limit: the maximum current that the side can carry.

[0068] Power loss: The amount of energy lost as electricity flows through a line.

[0069] Line Impedance: Factors such as resistance or inductance that affect the transmission of power.

[0070] Structure of the topological graph: The topological graph of the entire power system consists of nodes and edges, forming a directed graph, where each edge represents the direction and restriction of power flow.

[0071] After using simulation software to simulate and test the topology model, the objective function is used to calculate the target value of the topology model. The objective function expression is: , where is the target value, is the power system stability index, is the load deviation amplitude, is the power loss, , , are the power system stability index, load deviation amplitude and power loss adjustment coefficient respectively, and , , Both are greater than 0.

[0072] The calculation logic of the load deviation amplitude is: obtain the actual load of each load node in the current topology model, and calculate the load deviation of each load node. The expression is: , where is the load deviation, is the actual load, is the standard load. After obtaining the load deviation of each load node, the mean value of the load deviation and the standard deviation of the load deviation are calculated. The expression is: , where is the load deviation standard deviation, is the mean value of load deviation, For the The load deviation of each load node, is the number of load nodes, and the load deviation amplitude is obtained by dividing the load deviation mean by the load deviation standard deviation. The larger the load deviation amplitude is, the more unbalanced the overall load of the current topology model is; The calculation logic of the load deviation amplitude is very clear, reflecting the impact of load imbalance on power system optimization. In this way, the degree of load distribution imbalance can be quantified and provide guidance for optimization in the topology model. Specifically, the larger the load deviation amplitude, the more unbalanced the load distribution, and the power flow of the system may be subject to greater fluctuations, resulting in higher power loss, frequency fluctuations and other problems, affecting the stability and efficiency of the system.

[0073] Therefore, the load deviation amplitude, as an evaluation indicator, can help determine the optimization direction of the current topology model: Large load deviation amplitude: indicates that the system load is unevenly distributed, which may lead to problems such as unbalanced power flow and voltage fluctuation. At this time, the topology needs to be optimized, such as adjusting the location of load nodes or the allocation of power nodes, reconfiguring the power transmission path, etc.

[0074] The load deviation amplitude is small: it means that the load distribution is more balanced, the power system may be more stable and the power loss is lower, and the topology structure performs better.

[0075] By analyzing the load deviation amplitude, the topology of the power system can be effectively optimized, the impact of unbalanced loads can be reduced, and the overall stability and efficiency of the system can be improved.

[0076] The power loss calculation expression is: , where For Node and nodes The current value between For Node and nodes The resistance value of the line between them; Power loss refers to the energy loss caused by the resistance of the power line during power transmission. Power loss is closely related to the transmission path, current size and voltage difference in the topology diagram. The resistance and reactance of the transmission line determine the size of the power loss. Through the topology diagram, high-impedance lines and inefficient lines can be identified, power flow can be optimized, and losses can be reduced. The topology determines the direction of current flow and the distribution of voltage. By adjusting the topology (for example, selecting different power nodes or optimizing the power flow path), power loss can be effectively reduced. When optimizing the topology model, smaller power loss indicates that the topology model performs better.

[0077] The calculation logic of the power system stability index is: real-time monitoring of the output current fluctuation value of the distributed power generation node. When the output current fluctuation value is greater than the fluctuation threshold, the power system stability index is calculated. The expression is: , where is the power system stability index, , Node and nodes The voltage value, , For Node The admittance matrix elements between , Node and nodes The phase angle of In power system analysis, the node-to-node admittance matrix (Y-matrix) is the core matrix that describes the electrical connection relationship between nodes in the power network. When calculating the stability of the power system, the admittance matrix not only plays an important role in power flow calculation, but also provides important information in dynamic stability analysis; The elements in the admittance matrix reflect the electrical characteristics between nodes. The Y-matrix is ​​a symmetrical matrix whose element size determines the current transmission relationship between nodes. In the entire power system, a complete admittance matrix is ​​constructed by considering all lines, equipment and compensation devices.

[0078] In this application, the node The method for obtaining the elements of the admittance matrix between belongs to the prior art and will not be described in detail in this application.

[0079] The power system stability index reflects the stability and performance of the power system. The larger the stability index, the better the stability of the power system, that is, the system can better maintain power balance, frequency and voltage stability, and the power flow is more balanced.

[0080] The relationship between the fixed index and the performance of the topology model: The performance of the topology model is closely related to the stability of the system. By optimizing the topology (such as adjusting the power transmission path between nodes, selecting appropriate generators and load distribution), the stability of the system can be improved, power loss can be reduced, and the load balance of the system can be improved.

[0081] Improve stability index: Topology optimization of the power system can improve the stability of the power system and reduce voltage fluctuations, frequency deviations and power losses.

[0082] The relationship between the stability index and the target of topology optimization: When the stability index is high, it means that the topological model of the power system is more optimized, and the system is more resilient and stable under different load and power generation conditions, thereby reducing the need for further optimization.

[0083] Updating the optimal path between nodes through the path algorithm includes the following steps: Source Node: Select a starting node in the topology graph, usually a power generation node or a load node.

[0084] Target Node: Select another node in the topology diagram, usually the target load point or transformer node.

[0085] Initializing the Dijkstra algorithm requires setting up the following data structures: Distance: Records the shortest distance from the source node to each node (initially, the distance to the source node is 0, and the distances to other nodes are infinite).

[0086] Visited Set: used to record the nodes for which the shortest path has been found.

[0087] Predecessor: Records the predecessor node of each node for path tracing.

[0088] The core steps of executing Dijkstra's algorithm are as follows: Set the distance of the source node to 0.

[0089] Set the distances of all other nodes to infinity (indicating they have not yet been reached).

[0090] Starting from the current node, calculate the path from this node to all adjacent nodes (neighboring nodes).

[0091] For each neighbor node, calculate the distance from the source node to the neighbor node: if the calculated distance is less than the known distance, update the shortest path value of the neighbor node.

[0092] Among all the unvisited nodes, select the node that is closest to the source node (i.e. the node with the smallest distance).

[0093] Mark the node as visited and add it to the set of visited nodes.

[0094] Repeat the above process until all nodes have been visited or the shortest path to the target node has been found.

[0095] Path calculation: The shortest path from the source node to the target node is calculated using the Dijkstra algorithm. Whenever the shortest path of a node is updated, the path can be traced back through the predecessor node to eventually obtain the complete path from the source node to the target node.

[0096] In the power system, the path selection is not only the shortest path, but also needs to consider the following constraints: Current carrying capacity: Ensure that the current of the line does not exceed its maximum capacity. For each path update of the Dijkstra algorithm, the current carrying capacity of each edge in the path is checked.

[0097] Power loss: The power loss of each edge is proportional to the current through impedance calculation. The path with the least loss should be selected to optimize the efficiency of the power system.

[0098] Voltage stability: Path selection also needs to consider voltage stability to ensure that the voltage is stable between nodes.

[0099] Suppose we have a power system with the following four key nodes: Power generation node: A device responsible for generating electricity.

[0100] Transformer node: Converts electricity from one voltage level to another.

[0101] Load node 1: The first user point to which the power is transmitted.

[0102] Load node 2: The second user point to which the power is transmitted.

[0103] These four nodes are interconnected by several power transmission lines, each of which has a certain transmission capacity (e.g., current carrying capacity) and transmission impedance (affecting power loss). Our goal is to calculate the shortest path from the generation node to the load node through the Dijkstra algorithm, thereby optimizing power transmission and reducing losses and overloads.

[0104] Node and line settings: Node A: Power generation node Node B: Transformer node Node C: Load Node 1 Node D: Load Node 2 The lines connecting these nodes are as follows: Line 1: From the power generation node A to the transformer node B, the line has a large transmission capacity but a certain impedance.

[0105] Line 2: From transformer node B to load node 1, the line has a small transmission capacity and a large impedance.

[0106] Line 3: directly connected from power generation node A to load node 1, the impedance of the line is moderate.

[0107] Line 4: From transformer node B to load node 2, the line impedance is small.

[0108] First, we initialize the shortest path from each node to the source node (generation node A): The distance of the power generation node A is 0 (the distance from the starting point to itself is 0).

[0109] The distances of other nodes are initialized to "infinity" (indicating that the shortest path has not been found yet).

[0110] From all nodes, select the node with the shortest distance to the power generation node A (current shortest path). Initially, the distance to the power generation node A is 0, so we start from the A node.

[0111] Calculate the distance to node B: Connect via line 1 and record the path length.

[0112] Calculate the distance to node C: Connect via line 3 and record the path length.

[0113] Next, update the distances of the other nodes: For node B, the path from A to B via line 1 is shorter than the path from A to C, so node B is selected as the current node and the path is continued to be updated.

[0114] For node C, the path from A to C is already shorter via a direct connection and does not need to go through node B.

[0115] After updating the paths of nodes B and C, we continue to select the node with the shortest distance as the next node for path update. At this point, we have obtained the shortest path information.

[0116] Finally, we get the optimal path from power generation node A to load node C and load node D: The shortest path from A to C is directly connected via line 3.

[0117] The shortest path from A to D is via line 1 and line 4.

[0118] Through the Dijkstra algorithm, we successfully found the optimal power transmission path, ensuring that the power flow from the power generation node to the load node is both efficient and reduces power loss. At the same time, it avoids possible line overloads in the system and improves the reliability of the power system.

[0119] The current distribution of the edges is adjusted by the graph cut algorithm to reduce power loss and avoid overload, including the following steps: The basic idea of ​​the cut algorithm is to divide the network graph into two subgraphs to optimize certain objectives (such as power flow distribution). In power systems, graph cuts can be used to optimize current flow so that the load on the line does not exceed its capacity limit, reduce power loss, and ultimately optimize the stability and performance of the power system.

[0120] The graph cut algorithm transforms the topology of the power system into a flow network problem, where: Source node: Usually a power generation node or load point is selected as the source.

[0121] Sink: Select a load node or target terminal node as the sink.

[0122] Cut Sets: Control how current flows from sources to sinks by selecting cut sets in the graph.

[0123] CutSet: CutSet is the process of "cutting off" certain edges from the graph. By selecting appropriate edges to "cut" the power network, the flow of current can be controlled.

[0124] The core of the graph cut algorithm is to calculate the optimal cut set and adjust the current distribution. This is done through the following steps: Compute Flow for Cut Sets: For each cut set, compute the flow, the current flow from source to sink.

[0125] Finding the optimal cut set: The graph cut algorithm calculates the flow of different cut sets and finds the cut set that can reduce power loss and avoid overload.

[0126] Update current distribution: Adjust the current flow according to the optimal cut sets to ensure that the current distribution in the system meets the power flow constraints and line current limits.

[0127] When executing the graph cut algorithm, the current distribution needs to meet the system constraints, including: Current capacity constraint: The current flow of each line cannot exceed its maximum carrying capacity. The graph cut algorithm can ensure that the current flow of each edge does not exceed the current capacity of the line by calculating the optimal cut set.

[0128] Power loss constraint: The power loss in the power system is mainly related to the line impedance and current. Through the graph cut algorithm, the optimal current path is selected to reduce unnecessary power loss.

[0129] Suppose we have a topological diagram of a power system, which contains the following nodes and connecting lines: Power station (source node): Power generation equipment that provides electricity.

[0130] Substation (relay node): increases or decreases voltage and adjusts the current transmission path.

[0131] Two load points (terminal nodes): two different electricity users (such as factories, residential areas, etc.).

[0132] These nodes are connected by several transmission lines, each with different carrying capacity.

[0133] Our goal is to optimize the distribution of current through the graph cut algorithm to ensure that the current of each line does not exceed its carrying capacity while reducing power loss, thereby improving the stability and performance of the system.

[0134] First, the topology of the power system is converted into a flow network. The source node (power station) provides power, and the terminal node (load point) receives power. Each line in the diagram has a certain current carrying capacity and impedance.

[0135] The core of graph cuts is to select "cut sets", that is, to "cut off" certain lines (edges) in the graph, thereby separating the paths of power flow. By selecting appropriate cut sets, we can control the way current flows and optimize power transmission.

[0136] Cut set selection: We select several key routes from the power station to the load point as cut sets. When selecting cut sets, we need to consider the following factors: The current carrying capacity of each line.

[0137] Power loss per line.

[0138] For each selected cut set, calculate the flow rate of power flow. The flow rate is the current size when power flows from the source node (power station) through each line to the sink node (load point).

[0139] Calculating flow: We need to estimate the current flowing through each line to ensure that it does not exceed the maximum carrying capacity of the line. If a line is about to be overloaded, the graph cut algorithm will automatically adjust the current path and select other lines to share the load.

[0140] The graph cut algorithm compares the flow of different cut sets to find the optimal cut set that can reduce power loss and avoid overload. Specifically, the algorithm will find a set of edges (lines) that make the current flow more balanced, reduce unnecessary power loss, and ensure that the current of each line does not exceed its carrying capacity.

[0141] Optimal cut set selection: In actual operation, the optimal cut set may include multiple paths. The graph cut algorithm will select the most suitable solution by comparing the current flow of different paths.

[0142] Once the optimal cut set is found, the graph cut algorithm adjusts the distribution of current so that the current flows along the optimal path, avoiding overloading of certain lines in the system and minimizing power loss.

[0143] Adjust the current path: Based on the optimal cut set, the algorithm will redirect the current. For example, if the current of a line is close to the upper limit of the carrying capacity, the system will direct part of the current from this line to other lines to avoid overload.

[0144] example: Assume that a power station supplies power to two load points through three lines: The first line connects the power station and load point 1, and has a large current carrying capacity, but a high line impedance.

[0145] The second line connects the power station and load point 2. The line impedance is moderate and the current carrying capacity is small.

[0146] The third line connects the substation and load point 1, has a smaller current carrying capacity and a higher line impedance.

[0147] In actual power flow, there may be a situation where if a large amount of current flows to load point 1, the current of the first line will approach or exceed its carrying capacity. At this time, the graph cut algorithm will choose to transmit part of the current through the second line (connecting the power station and load point 2) or the third line (forwarded to load point 1 through the substation) to avoid overload and reduce power loss.

[0148] Through the optimization of the graph cut algorithm, the current flow in the power system is more balanced and the load of each line is reasonably distributed. The system not only avoids overload, but also successfully reduces power loss during power transmission, improving the stability and overall performance of the system.

[0149] The graph cut algorithm ensures the reasonable current distribution of the power system by intelligently selecting and adjusting the current path, avoiding current overload, reducing power loss, and improving the efficiency and stability of the power system. This method is particularly suitable for complex power systems and dealing with flow and constraint problems in power networks.

[0150] Repeat the above steps for multiple iterations. When the convergence conditions are met, the final topology model is output as the comprehensive load model, including the following steps: Repeat the target value calculation, optimal path update and current distribution adjustment steps for the current topology model for multiple iterations. When the number of iterations is equal to the number threshold or the target value of the topology model after a certain update is greater than or equal to the performance threshold, it is determined that the convergence condition is met.

[0151] When the convergence condition is met, the topology model obtained in each iteration is output, and the topology models obtained in each iteration are sorted from large to small according to the target value, and a model sorting table is obtained. The topology model ranked first in the model sorting table is selected as the final topology model, and the final topology model is output as the comprehensive load model; The comprehensive load model is applied in the distribution network. After recording the target value of the comprehensive load model at multiple time points, it is analyzed whether the comprehensive load model needs to be replaced. If it needs to be replaced, the corresponding comprehensive load model is selected according to the model sorting table.

[0152] After applying the comprehensive load model to the distribution network and recording the target values ​​of the comprehensive load model at multiple time points, analyzing whether the comprehensive load model needs to be replaced includes the following steps: After obtaining the target values ​​of the comprehensive load model at multiple time points, the maximum target value is selected as the target peak value, the target deviation is calculated based on the performance threshold and the target peak value, the target values ​​at multiple time points are integrated to obtain the target cumulative index, and the target values ​​at multiple time points are analyzed to obtain the target fluctuation amplitude. After the target deviation, target cumulative index and target fluctuation amplitude are amplified, the selection coefficient of the comprehensive load model is obtained by comprehensive calculation, and the expression is: , where is the selection coefficient, is the target deviation, is the target cumulative index, is the target fluctuation amplitude, , , are the information gains of target deviation, target cumulative index and target fluctuation amplitude respectively, and , , All are greater than 0; The smaller the selection coefficient, the worse the overall performance of the comprehensive load model applied in the distribution network, that is, the comprehensive load model needs to be replaced to ensure the operating stability and efficiency of the distribution network. The obtained selection coefficient is compared with the preset selection threshold. The selection threshold is used to determine whether the comprehensive load model needs to be replaced. If the selection coefficient is greater than or equal to the selection threshold, the overall performance of the comprehensive load model applied in the distribution network is good, and it is determined that the comprehensive load model does not need to be replaced. If the selection coefficient is less than the selection threshold, the overall performance of the comprehensive load model applied in the distribution network is poor, and it is determined that the comprehensive load model needs to be replaced.

[0153] When it is determined that the comprehensive load model needs to be replaced, the comprehensive load models are selected in order according to the model list, and the comprehensive load models in use are regularly analyzed.

[0154] The target deviation is calculated based on the performance threshold and the target peak. The calculation logic is: the target deviation is obtained by subtracting the target peak from the performance threshold. The larger the target deviation, the worse the application performance of the comprehensive load model. The target cumulative index is obtained by integrating the target values ​​at multiple time points. The expression is: , where represents the operation time of the comprehensive load model in the distribution network, It represents the target value of the comprehensive load model at a certain moment in time. The larger the target cumulative index is, the better the cumulative performance of the comprehensive load model in the distribution network process is.

[0155] After analyzing the target values ​​at multiple time points, the target fluctuation amplitude is obtained. The expression is: , where is the number of time points, For the The target value at a time point, is the target mean value. The larger the target fluctuation amplitude is, the worse the stability of the comprehensive load model is when applied in the distribution network.

[0156] After obtaining the target values ​​of the comprehensive load model at multiple time points, the application selects the maximum target value as the target peak value, calculates the target deviation based on the performance threshold and the target peak value, integrates the target values ​​at multiple time points to obtain the target cumulative index, and analyzes the target values ​​at multiple time points to obtain the target fluctuation amplitude. After gaining the target deviation, the target cumulative index and the target fluctuation amplitude, the selection coefficient of the comprehensive load model is obtained by comprehensive calculation. By comprehensively analyzing the application status of the comprehensive load model at multiple time points, it is determined whether the comprehensive load model needs to be replaced, thereby further ensuring the stability of the distribution network operation and improving the efficiency.

[0157] Embodiment 3: A system for establishing a comprehensive load model containing distributed power sources described in this embodiment includes a topology model establishment module, an optimization module, and an iterative output module; Topology model building module: After obtaining the power system information through the API interface, a topology model of the power system is established, and each component in the power system (load, distributed power source, transmission line, etc.) is abstracted into nodes and edges in the topology model. Each node represents a load point or distributed power source point in the power system, and the edge represents the power transmission relationship between nodes. Node definition: Each node can be a load point (such as a home, factory, etc.) or a distributed power source (such as photovoltaic, wind power, etc.). The node can store its load value, power generation, historical data and other information. Edge definition: The edge represents the power transmission path, connecting different nodes. The weight of the edge can represent the transmission capacity of the line, current limit, etc. The topology model is sent to the optimization module; Optimization module: After using simulation software to simulate and detect the topology model, the target value of the topology model is calculated using the objective function, the optimal path between nodes is updated through the path algorithm, and the current distribution of the edge is adjusted through the graph cut algorithm; Iterative output module: Repeat the steps in the optimization module for multiple iterations. When the convergence conditions are met, the final topology model is output as the comprehensive load model.

[0158] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0159] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0160] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for establishing a comprehensive load model containing distributed power sources, characterized in that: The establishment method comprises the following steps: S1: After the system is established to obtain power system information through the API interface, a topological graph model of the power system is established, and each component in the power system is abstracted into nodes and edges in the topological graph model; S2: After using simulation software to simulate and test the topology model, the target value of the topology model is calculated using the objective function, the optimal path between nodes is updated through the path algorithm, and the current distribution of the edge is adjusted through the graph cut algorithm; S3: Repeat step S2 for multiple iterations, and when the convergence condition is met, output the final topology model as the comprehensive load model; The objective function is used to calculate the target value of the topology model. The objective function expression is: , where is the target value, is the power system stability index, is the load deviation amplitude, is the power loss, , , are the power system stability index, load deviation amplitude and power loss adjustment coefficient respectively, and , , Both are greater than 0.

2. The method for establishing a comprehensive load model containing distributed power sources according to claim 1, characterized in that: Repeat step S2 for multiple iterations to determine whether the convergence condition is met, including the following steps: Repeat the target value calculation, optimal path update and current distribution adjustment steps for the current topology model for multiple iterations. When the number of iterations is equal to the number threshold or the target value of the topology model after a certain update is greater than or equal to the performance threshold, it is determined that the convergence condition is met.

3. The method for establishing a comprehensive load model containing distributed power sources according to claim 2, characterized in that: When the convergence conditions are met, the final topology model is output as a comprehensive load model, including the following steps: When the convergence condition is met, the topology model obtained in each iteration is output, and the topology models obtained in each iteration are sorted from large to small according to the target value, and a model sorting table is obtained. The topology model ranked first in the model sorting table is selected as the final topology model, and the final topology model is output as the comprehensive load model; The comprehensive load model is applied in the distribution network. After recording the target value of the comprehensive load model at multiple time points, it is analyzed whether the comprehensive load model needs to be replaced. If it needs to be replaced, the corresponding comprehensive load model is selected according to the model sorting table.

4. The method for establishing a comprehensive load model containing distributed power sources according to claim 3, characterized in that: The power loss calculation expression is: , where For Node and nodes The current value between For Node and nodes The resistance value of the line between them; The calculation logic of the power system stability index is: real-time monitoring of the output current fluctuation value of the distributed power supply node, when the output current fluctuation value is greater than the fluctuation threshold, the power system stability index is calculated, and the expression is: , where is the power system stability index, , Node and nodes The voltage value, , For Node The admittance matrix elements between , Node and nodes The phase angle of .

5. The method for establishing a comprehensive load model containing distributed power sources according to claim 4, characterized in that: The calculation logic of the load deviation amplitude is: obtain the actual load of each load node in the current topology model, and calculate the load deviation of each load node. The expression is: , where is the load deviation, is the actual load, is the standard load. After obtaining the load deviation of each load node, the mean value of the load deviation and the standard deviation of the load deviation are calculated. The expression is: , where is the load deviation standard deviation, is the mean value of load deviation, For the The load deviation of each load node, is the number of load nodes, and the load deviation amplitude is obtained by dividing the load deviation mean by the load deviation standard deviation.

6. The method for establishing a comprehensive load model containing distributed power sources according to claim 5, characterized in that: After the system is established to obtain power system information through the API interface, a topological graph model of the power system is established, and each component in the power system is abstracted into nodes and edges in the topological graph model, including the following steps: Establish a system to obtain various information of the power system in real time through the API interface, including demand data of load points, power generation data of distributed power sources, data of power transmission lines, and operating status of the power system; Each node represents a load point in the power system, including residential load points, industrial load points, commercial load points, and factory load points. The attributes of each load point include load value, historical data, distributed power type, power generation, power generation historical data, and power generation mode. The edges represent the power transmission paths connecting different nodes, connecting the load points and the power supply points, and each edge represents the power flow relationship between the nodes; The topological graph model of the power system consists of nodes and edges, forming a directed graph, where each edge represents the direction and restriction of power flow.

7. The method for establishing a comprehensive load model containing distributed power sources according to claim 3, characterized in that: After applying the comprehensive load model to the distribution network and recording the target values ​​of the comprehensive load model at multiple time points, analyzing whether the comprehensive load model needs to be replaced includes the following steps: After obtaining the target values ​​of the comprehensive load model at multiple time points, the maximum target value is selected as the target peak value, the target deviation is calculated based on the performance threshold and the target peak value, the target values ​​at multiple time points are integrated to obtain the target cumulative index, and the target values ​​at multiple time points are analyzed to obtain the target fluctuation amplitude. After the target deviation, target cumulative index and target fluctuation amplitude are amplified, the selection coefficient of the comprehensive load model is obtained by comprehensive calculation, and the expression is: , where is the selection coefficient, is the target deviation, is the target cumulative index, is the target fluctuation amplitude, , , are the information gains of target deviation, target cumulative index and target fluctuation amplitude respectively, and , , All are greater than 0; The obtained selection coefficient is compared with the preset selection threshold. The selection threshold is used to determine whether the comprehensive load model needs to be replaced. If the selection coefficient is greater than or equal to the selection threshold, the overall performance of the comprehensive load model applied in the distribution network is analyzed to be good, and it is determined that the comprehensive load model does not need to be replaced. If the selection coefficient is less than the selection threshold, the overall performance of the comprehensive load model applied in the distribution network is analyzed to be poor, and it is determined that the comprehensive load model needs to be replaced.

8. The method for establishing a comprehensive load model containing distributed power sources according to claim 7, characterized in that: The target deviation is calculated based on the performance threshold and the target peak. The calculation logic is: the target deviation is obtained by subtracting the target peak from the performance threshold; The target cumulative index is obtained by integrating the target values ​​at multiple time points. The expression is: , where represents the operation time of the comprehensive load model in the distribution network, represents the target value of the comprehensive load model at the time instant; After analyzing the target values ​​at multiple time points, the target fluctuation amplitude is obtained. The expression is: , where is the number of time points, For the The target value at a time point, is the target mean.

9. A system for establishing a comprehensive load model containing distributed power sources, used to implement the establishment method according to any one of claims 1 to 8, characterized in that: It includes topology model building module, optimization module and iterative output module; Topology model building module: After obtaining power system information through the API interface, a topology model of the power system is established, and each component in the power system is abstracted into nodes and edges in the topology model; Optimization module: After using simulation software to simulate and detect the topology model, the target value of the topology model is calculated using the objective function, the optimal path between nodes is updated through the path algorithm, and the current distribution of the edge is adjusted through the graph cut algorithm; Iterative output module: Repeat the steps in the optimization module for multiple iterations. When the convergence conditions are met, the final topology model is output as the comprehensive load model.

Images