Intelligent power distribution network accurate topology identification method based on complex asymmetric game

Abstract

The invention discloses an intelligent power distribution network accurate topology identification method based on a complex asymmetric game, and relates to the technical field of power distribution network accurate topology identification. The identification method comprises the following steps: data acquisition: parameter acquisition of a transformer, a JP cabinet and a user meter box in a power distribution network is involved; the data is used for analyzing relevance, load balancing and hierarchical priority among the equipment, the data comprises electrical parameters, geographical location information, load data and equipment characteristic weight, and the data is collected through the intelligent electric meter, the power distribution terminal and the remote monitoring equipment; data preprocessing: the acquired original data has the problems of missing values and abnormal values; according to the precise topology identification method for the intelligent power distribution network based on the complex asymmetric game, simulation data are generated, a complex asymmetric game model is used for being compared with a traditional topology identification method, and the accuracy and effectiveness of the complex asymmetric game model in hierarchical decision making and topology optimization are verified.

Description

Technical Field

[0001] The present invention relates to the technical field of precise topology identification of distribution networks, and in particular to a precise topology identification method for intelligent distribution networks based on complex asymmetric games. Background Art

[0002] The power distribution network is an important part of the power system, and its topology directly affects the operation efficiency and reliability of the power grid. In actual operation, the distribution network usually has a hierarchical three-level structure, including transformers, JP cabinets and user meter boxes. How to accurately identify and optimize this three-level topology is one of the core issues in the planning and operation of the distribution network.

[0003] At present, the distribution network topology identification mostly adopts static modeling or optimization methods based on a single indicator. These methods have the following shortcomings:

[0004] Lack of dynamics: Traditional methods are usually based on a fixed grid topology model, which makes it difficult to adapt to changes in load distribution or equipment status in the distribution network in real time.

[0005] Single factor optimization: Many methods only consider a single factor (such as the correlation of electrical parameters between devices), ignoring other key factors such as layer priority and load balancing, resulting in insufficient accuracy in topology identification.

[0006] Insufficient complexity: In a multi-layer structure, simple linear optimization or correlation analysis cannot fully reflect the complex interactions between devices.

[0007] Therefore, in view of this, the existing structure and deficiencies are studied and improved, and a precise topology identification method for smart distribution networks based on complex asymmetric games is proposed. Summary of the invention

[0008] In view of the deficiencies in the prior art, the present invention provides a method for accurately identifying the topology of an intelligent distribution network based on complex asymmetric games, which solves the problems raised in the above-mentioned background technology.

[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for accurately identifying the topology of a smart distribution network based on complex asymmetric games, the identification method comprising the following steps:

[0010] Data collection: involves the collection of parameters of transformers, JP cabinets and user meter boxes in the distribution network. The data will be used to analyze the correlation, load balancing and hierarchical priority between devices. The data includes electrical parameters, geographic location information, load data, and device characteristic weights. The data is collected through smart meters, distribution terminals and remote monitoring equipment.

[0011] Data preprocessing: The collected raw data has missing values ​​and outliers, which need to be cleaned by the following methods: Missing value filling: Use historical mean values ​​to fill in missing load or electrical parameters. In order to eliminate the dimension difference of the data, all parameters are standardized using mean-variance standardization; the cleaned electrical parameters, geographic location information, load data and equipment characteristic weights are combined into a unified data matrix X for subsequent calculations;

[0012] Profit calculation of complex asymmetric game models: The profit value of each device is calculated through the multi-factor profit function Ui, and the strategy is dynamically adjusted and a penalty mechanism is introduced during the profit calculation process to optimize the hierarchical allocation of devices;

[0013] Hierarchical decision-making and topology optimization module: According to the calculation results of equipment benefits UiU_iUi, hierarchical decisions are made on the equipment. Transformer layer: high-benefit equipment is allocated to give priority to the core operation of the power grid. JP cabinet layer: the second highest-benefit equipment is allocated to share the load of the transformer. User meter box layer: the remaining equipment is allocated to the user meter box layer.

[0014] Furthermore, electrical parameters include current, voltage, and power, which are used to measure the electrical relationships between devices.

[0015] Furthermore, the geographic location information is the latitude and longitude coordinates of the device, which is used to analyze the spatial distribution and distance correlation of the device.

[0016] Furthermore, the load data includes the current load and historical load of each device, which is used for load balancing calculations.

[0017] Furthermore, the device characteristic weights are hierarchical priority: different priority weights are set according to the hierarchical level of the device, and importance weight: weights are assigned according to the criticality of the device.

[0018] Furthermore, the data collection method uses one or more of the power communication protocol, geographic information system, historical database, real-time data, and static data.

[0019] Furthermore, the devices are initially allocated according to the tier capacity and device type:

[0020] Transformer layer: the device with the highest yield is allocated;

[0021] JP cabinet layer: The second highest-yield equipment is allocated to the JP cabinet layer;

[0022] User meter box layer: The remaining devices are allocated to the user meter box layer.

[0023] Furthermore, the equipment benefit value Ui is the primary decision-making basis, and the higher the benefit, the higher the priority.

[0024] Further, intra-layer optimization: within each layer, in order to improve operational efficiency and equipment synergy, equipment arrangement and allocation are further optimized.

[0025] Furthermore, the core of topology optimization is to ensure that the connection relationship between devices in the hierarchical structure is reasonable and the load is balanced, and to improve the overall performance of the hierarchical topology through dynamic adjustment.

[0026] The present invention provides a precise topology identification method for an intelligent distribution network based on complex asymmetric games, which has the following beneficial effects: by generating simulation data, using complex asymmetric game models and traditional topology identification methods for comparison, the accuracy and effectiveness of the complex asymmetric game model in hierarchical decision-making and topology optimization are verified, the load balancing and correlation matching of the hierarchical structure are improved, the hierarchical stability of high-priority devices is ensured, and it is more suitable for complex hierarchical topology identification tasks than traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a flow chart of the data preprocessing part of the present invention;

[0028] Figure 2 This is a hierarchical decision-making flow chart of the present invention. DETAILED DESCRIPTION

[0029] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0030] like Figure 1 , Figure 2 As shown, the present invention provides a technical solution: a method for accurately identifying the topology of a smart distribution network based on complex asymmetric games, the identification method comprising the following steps:

[0031] Data collection: involves the collection of parameters of transformers, JP cabinets and user meter boxes in the distribution network. The data will be used to analyze the correlation, load balancing and hierarchical priority between devices. The data includes electrical parameters, geographic location information, load data, and device characteristic weights. The data is collected through smart meters, distribution terminals and remote monitoring equipment.

[0032] Data preprocessing: The collected raw data has missing values ​​and outliers, which need to be cleaned by the following methods: Missing value filling: Use historical mean values ​​to fill in missing load or electrical parameters. In order to eliminate the dimension difference of the data, all parameters are standardized using mean-variance standardization; the cleaned electrical parameters, geographic location information, load data and equipment characteristic weights are combined into a unified data matrix X for subsequent calculations;

[0033] Profit calculation of complex asymmetric game models: The profit value of each device is calculated through the multi-factor profit function Ui, and the strategy is dynamically adjusted and a penalty mechanism is introduced during the profit calculation process to optimize the hierarchical allocation of devices;

[0034] Hierarchical decision-making and topology optimization module: According to the calculation results of equipment benefits UiU_iUi, hierarchical decisions are made on the equipment. Transformer layer: high-benefit equipment is allocated to give priority to the core operation of the power grid. JP cabinet layer: the second highest-benefit equipment is allocated to share the load of the transformer. User meter box layer: the remaining equipment is allocated to the user meter box layer.

[0035] Electrical parameters include current, voltage, and power, which are used to measure the electrical relationships between devices.

[0036] The geographic location information is the latitude and longitude coordinates of the device, which is used to analyze the spatial distribution and distance correlation of the device.

[0037] Load data includes the current load and historical load of each device and is used for load balancing calculations.

[0038] The device characteristic weights are hierarchical priority: different priority weights are set according to the level of the device, and importance weight: weights are assigned according to the criticality of the device.

[0039] The data collection method uses one or more of the following: power communication protocol, geographic information system, historical database, real-time data, and static data.

[0040] Initially allocate devices according to tier capacity and device type:

[0041] Transformer layer: the device with the highest yield is allocated;

[0042] JP cabinet layer: The second highest-yield equipment is allocated to the JP cabinet layer;

[0043] User meter box layer: The remaining devices are allocated to the user meter box layer.

[0044] The equipment benefit value Ui is the primary decision-making basis. The higher the benefit, the higher the priority.

[0045] Intra-layer optimization: Within each layer, equipment arrangement and allocation are further optimized to improve operational efficiency and equipment synergy.

[0046] The core of topology optimization is to ensure that the connection relationship between devices in the hierarchical structure is reasonable and the load is balanced, and to improve the overall performance of the hierarchical topology through dynamic adjustment.

[0047] The collected raw data may contain missing values, outliers, and other problems, and need to be cleaned by the following methods:

[0048] Missing value filling:

[0049] Use historical averages to fill in missing load or electrical parameters.

[0050] If the missing rate exceeds 30%, the data of the device will be discarded.

[0051] Outlier Detection:

[0052] Detect outliers using the 3 times standard deviation method:

[0053] X i ∈[μ-3σ,μ+3σ]

[0054] Outliers are replaced with historical medians.

[0055] Data Standardization

[0056] To eliminate the dimension difference of the data, all parameters were standardized using mean-variance standardization.

[0057] formula:

[0058]

[0059] X: original data value.

[0060] μ: data mean.

[0061] σ: standard deviation of data.

[0062] Data dimension combination

[0063] The cleaned electrical parameters, geographic location information, load data and equipment characteristic weights are combined into a unified data matrix X for subsequent calculations.

[0064] Data matrix structure

[0065]

[0066] Each row corresponds to a device, and each column corresponds to a characteristic of the device.

[0067] The core of the complex asymmetric game model is to calculate the benefit value of each device through the multi-factor benefit function Ui, and dynamically adjust the strategy and introduce a penalty mechanism during the benefit calculation process to optimize the hierarchical allocation of devices. The following is a detailed description of the game process, benefit calculation process, strategy adjustment process and penalty mechanism, with detailed formulas and examples:

[0068] In the distribution network topology identification, transformers, JP cabinets, and user meter boxes belong to different levels and have an asymmetric relationship with each other: the transformer has the highest priority and affects the allocation of lower-level equipment. The interaction between the JP cabinet and the user meter box is mainly affected by the geographical location and load. After the user meter box is allocated to the JP cabinet, it has a local competitive relationship with other user meter boxes. The asymmetric game model comprehensively considers these factors, calculates the benefits of each device and adjusts its allocation strategy.

[0069] Participants: NNN devices, numbered i=1,2,...,N

[0070] Strategy space: The strategy of each device is its hierarchical selection Si, that is, the choice of allocation to the transformer, JP cabinet or user meter box.

[0071] Profit function: The profit Ui of each device depends on the association between devices, hierarchical priority, load balancing and importance weight.

[0072] Detailed calculation process of the profit function

[0073] Profit function formula

[0074] The total profit function consists of the following four parts:

[0075] U i =U 关联性 +U 层级 +U 负载 +U 重要性

[0076] Itemized income formula

[0077] Relevance benefit: reflects the relevance between the device and its potential distribution target:

[0078]

[0079] R ij : The association between devices i and j.

[0080] w j : Hierarchy priority of the target device.

[0081] Tier priority benefits: Different basic benefits are given according to the priority of the tier to which the device belongs

[0082]

[0083] L i , L j : Load of device i and target device j.

[0084] L max :Measure the maximum load value.

[0085] Importance Weighted Benefit: The importance of a device directly affects its benefit:

[0086] U 重要性 =δ·I i

[0087] I i : Importance weight of the device.

[0088] Strategy Adjustment Process

[0089] Policy Definition

[0090] The policy Si of each device i is its hierarchical choice. For a three-level topology, the policy space is:

[0091] S i ∈{transformer layer, JP cabinet layer, user meter box layer}

[0092] Dynamic adjustment rules

[0093] Initial Strategy: All devices are initially tiered based on the highest priority and relevance target assignment.

[0094] Dynamic update of benefits: After each round of game, the benefit function UiU_iUi is calculated according to the current strategy, and the hierarchical selection of devices is dynamically adjusted.

[0095] Strategy convergence condition: If the strategy of the device in two rounds of game no longer changes, the game converges.

[0096] formula:

[0097]

[0098] The strategy of device i in round t+1.

[0099] The payoff of device i in round t.

[0100] Penalty mechanism:

[0101] Penalty target: In order to prevent devices from always choosing a specific layer (such as too many devices concentrated in the transformer layer), a penalty mechanism is introduced to adjust the benefits of high-load or low-correlation devices.

[0102] Penalty formula: Overload penalty:

[0103] If the load on a device exceeds the capacity of its target tier, revenue is reduced:

[0104]

[0105] C 目标层 : The capacity of the target layer.

[0106] Penalty for insufficient relevance:

[0107] If the device's relevance to the target layer is below the threshold, the revenue is reduced:

[0108]

[0109] Relevance threshold.

[0110] Total Revenue Adjustment Formula

[0111] Total benefits after introducing the penalty mechanism:

[0112] U′ i =U i +P 负载 +P 关联性

[0113] Detailed description of the game process

[0114] Initialization phase:

[0115] Initialization:

[0116] All equipment is divided into three levels (transformer, JP cabinet, user meter box), and the equipment attributes include: electrical parameters (such as current, voltage, power); load information; level priority and importance weight.

[0117] Initialize the device policy set

[0118] S i ∈{transformer layer, JP cabinet layer, user meter box layer} The initial strategy is to give priority to the layer with the greatest correlation.

[0119] Initial profit calculation:

[0120] Calculate the initial revenue of each device according to the formula

[0121]

[0122] Store the calculation results as an initial income table.

[0123] Game Iteration Process

[0124] In each iteration, all devices adjust their stratification according to the current strategy and update their revenue values ​​until the strategy converges. The specific steps are as follows:

[0125] Earnings Update

[0126] For each device i, recalculate its revenue

[0127] Consider the current strategy and the status of associated devices to calculate the benefits of each hierarchical target.

[0128] Updated income statement

[0129] formula:

[0130]

[0131] P 负载 : Negative load exceeding limit penalty.

[0132] P 关联性 : Penalty for insufficient relevance.

[0133] Strategy Adjustment

[0134] Based on the latest revenue per device Choose the strategy with the highest return

[0135] Policy adjustment rules:

[0136]

[0137] Adjust logic:

[0138] 1. Traverse each device and calculate its revenue under different layers.

[0139] 2. Select the layer with the highest return as the strategy for the next round.

[0140] The penalty mechanism introduces penalties for high-load or low-association devices and updates the benefits.

[0141] Punishment mechanism formula:

[0142] 1. Penalty for overload

[0143]

[0144] C 目标层 : Target layer capacity.

[0145] L i : Load of device i.

[0146] Insufficient relevance penalty

[0147]

[0148] Relevance threshold.

[0149] R ij : The association between device i and target device j.

[0150] The profit update formula after punishment is: U′i =U i +P 负载 +P 关联性

[0151] Convergence judgment

[0152] Conditions for judging whether the device policy is convergent:

[0153]

[0154] If the strategies of all devices do not change, the game ends.

[0155] Otherwise, proceed to the next iteration.

[0156] Game over

[0157] When the strategies of all devices converge, the game process ends and the final hierarchical structure is output.

[0158] Hierarchical decision making and topology optimization module

[0159] Overall goal of hierarchical decision making

[0160] Layering rules:

[0161] According to the calculation results of equipment benefits UiU_iUi, hierarchical decisions are made for equipment:

[0162] 1. Transformer layer: allocate high-yield equipment to prioritize the core operation of the power grid.

[0163] 2. JP cabinet layer: Allocate the second highest benefit equipment to share the load of the transformer.

[0164] 3. User meter box layer: The remaining devices are allocated to the user meter box layer.

[0165] Optimization goal: Ensure that the device allocation of each layer meets the following requirements: Maximize correlation: The correlation between devices within a layer is high, and the correlation between devices between layers is low. Load balancing: The load of devices on each layer is as balanced as possible to avoid overload or inefficient operation. Layer stability: High-priority devices are allocated to the core layer first.

[0166] Decision-making process

[0167] Initialize Layer

[0168] Profit sorting: Sort the devices in descending order according to the profit function Ui to form a profit list:

[0169] Sorting result = {U1, U2, ..., U N}, U1>U2>…>U N

[0170] The sorting results are used to guide the priority allocation of equipment.

[0171] Tiered initial allocation:

[0172] Initially allocate devices according to tier capacity and device type:

[0173] Transformer layer: Assign the highest gain devices.

[0174] JP cabinet layer: The second highest yield equipment is allocated to the JP cabinet layer.

[0175] User meter box layer: The remaining devices are allocated to the user meter box layer.

[0176] Decision Metrics: During the tiering process, the allocation decision for each device is based on the following metrics:

[0177] Benefit ranking: The equipment benefit value Ui is the primary decision-making basis. The higher the benefit, the higher the priority.

[0178] Tier capacity: The total load of devices on each tier cannot exceed the maximum capacity of that tier:

[0179]

[0180] If the capacity of the current layer is insufficient, the device will be allocated to the next layer.

[0181] Relevance threshold: The relevance between the device and the target layer must meet the minimum threshold R ij ≥R min , otherwise reallocate.

[0182] Load balancing: The load of the devices in each layer should be balanced as much as possible to avoid overloading a single device:

[0183]

[0184] Intra-layer optimization: In each layer, in order to improve operation efficiency and equipment synergy, further optimize equipment arrangement and allocation:

[0185] Relevance optimization: Rearrange the devices in the current layer to maximize the relevance between the devices:

[0186]

[0187] Prioritize grouping highly correlated devices together.

[0188] Load adjustment: adjust the load of the current layer equipment to make the load as balanced as possible

[0189]

[0190] L 目标 : Target load within the layer.

[0191] Topology optimization: The core of topology optimization is to ensure that the connection relationship between devices in the hierarchical structure is reasonable and the load is balanced, and to improve the overall performance of the hierarchical topology through dynamic adjustment.

[0192] Topology Optimization Goal

[0193] Inter-layer collaboration: The upper-layer equipment must be able to support the load requirements of the lower-layer equipment to form a stable power supply link.

[0194] Dynamic adjustment: When the load changes or the correlation changes, the tiers are adjusted dynamically.

[0195] Optimization Algorithm

[0196] Dynamic adjustment based on the relevance matrix: According to the relevance matrix RRR, the connection relationship between devices in different layers is optimized:

[0197]

[0198] Load transfer: When the load on a certain layer exceeds the limit, the load of some equipment will be transferred to the adjacent layer:

[0199]

[0200] Redistribution strategy: Redistribute the device tiers according to the updated benefit value Ui

[0201]

[0202] Convergence judgment: Convergence conditions for topology optimization:

[0203] Load balancing across all tiers reaches the threshold:

[0204]

[0205] The correlation change between devices at each layer does not exceed the threshold:

[0206] |ΔR ij |≤δ

[0207] This embodiment generates simulation data and compares the complex asymmetric game model with the traditional topology identification method to verify the accuracy and effectiveness of the complex asymmetric game model in hierarchical decision-making and topology optimization.

[0208] The embodiments of the present invention are given for the purpose of illustration and description, and are not intended to be exhaustive or to limit the invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

Claims

1. A method for accurate topology identification of smart distribution network based on complex asymmetric game, characterized by: The identification method comprises the following steps: Data collection: involves the collection of parameters of transformers, JP cabinets and user meter boxes in the distribution network. The data will be used to analyze the correlation, load balancing and hierarchical priority between devices. The data includes electrical parameters, geographic location information, load data, and device characteristic weights. The data is collected through smart meters, distribution terminals and remote monitoring equipment. Data preprocessing: The collected raw data has missing values ​​and outliers, which need to be cleaned by the following methods: Missing value filling: Use historical mean values ​​to fill in missing load or electrical parameters. In order to eliminate the dimension difference of the data, all parameters are standardized using mean-variance standardization; the cleaned electrical parameters, geographic location information, load data and equipment characteristic weights are combined into a unified data matrix X for subsequent calculations; Profit calculation of complex asymmetric game models: The profit value of each device is calculated through the multi-factor profit function Ui, and the strategy is dynamically adjusted and a penalty mechanism is introduced during the profit calculation process to optimize the hierarchical allocation of devices; Hierarchical decision-making and topology optimization module: According to the calculation results of equipment benefits UiU_iUi, hierarchical decisions are made on the equipment. Transformer layer: high-benefit equipment is allocated to give priority to the core operation of the power grid. JP cabinet layer: the second highest-benefit equipment is allocated to share the load of the transformer. User meter box layer: the remaining equipment is allocated to the user meter box layer.

2. According to claim 1, a method for accurately identifying topology of intelligent distribution network based on complex asymmetric game, characterized in that: Electrical parameters include current, voltage, and power, which are used to measure the electrical relationships between devices.

3. According to claim 1, a method for accurately identifying topology of intelligent distribution network based on complex asymmetric game, characterized in that: The geographic location information is the latitude and longitude coordinates of the device, which is used to analyze the spatial distribution and distance correlation of the device.

4. According to claim 1, a method for accurately identifying topology of intelligent distribution network based on complex asymmetric game, characterized in that: Load data includes the current load and historical load of each device and is used for load balancing calculations.

5. According to claim 4, a method for accurately identifying topology of intelligent distribution network based on complex asymmetric game, characterized in that: The device characteristic weights are hierarchical priority: different priority weights are set according to the level of the device, and importance weight: weights are assigned according to the criticality of the device.

6. According to claim 5, a method for accurately identifying topology of intelligent distribution network based on complex asymmetric game, characterized in that: The data collection method uses one or more of the following: power communication protocol, geographic information system, historical database, real-time data, and static data.

7. The method for accurately identifying topology of a smart distribution network based on complex asymmetric game according to claim 6 is characterized by: Initially allocate devices according to tier capacity and device type: Transformer layer: the device with the highest yield is allocated; JP cabinet layer: The second highest-yield equipment is allocated to the JP cabinet layer; User meter box layer: The remaining devices are allocated to the user meter box layer.

8. According to claim 1, a method for accurately identifying topology of intelligent distribution network based on complex asymmetric game, characterized in that: The equipment benefit value Ui is the primary decision-making basis. The higher the benefit, the higher the priority.

9. The method for accurately identifying topology of a smart distribution network based on complex asymmetric game according to claim 8, characterized in that: Intra-layer optimization: Within each layer, equipment arrangement and allocation are further optimized to improve operational efficiency and equipment synergy.

10. The method for accurately identifying topology of a smart distribution network based on complex asymmetric game according to claim 9, characterized in that: The core of topology optimization is to ensure that the connection relationship between devices in the hierarchical structure is reasonable and the load is balanced, and to improve the overall performance of the hierarchical topology through dynamic adjustment.

Images