Power distribution network structure form planning method and system considering multi-agent game

Abstract

The invention provides a power distribution network structure form planning method and system considering a multi-agent game, and relates to the technical field of power distribution network planning. Comprising the steps of building a charging station operator model based on charging station electricity selling income, electricity purchasing cost, operation and maintenance cost and a charging congestion penalty term by considering maximum income of a charging station operator; establishing a power distribution network operator model by considering the minimum cost of a power distribution network operator; establishing an EV vehicle owner decision objective function by considering that the charging cost and the time delay cost are minimum; solving a game equilibrium state by taking an EV owner as a subordinate participant of the game and taking a decision of a power distribution network operator and a charging station operator as a game main body; and determining a to-be-built line and structure of the power distribution network based on the game equilibrium state. According to the method, the multi-agent game behavior in the market is accurately simulated, the power distribution network operator strategy, the charging station operator strategy and the EV vehicle owner benefit conflict are effectively coordinated, it is ensured that each market agent continuously optimizes the decision in the game process, and the benefit maximization is achieved.

Description

Technical Field

[0001] This invention belongs to the field of power distribution network planning technology, and in particular relates to a method and system for planning the structure and form of power distribution networks that considers multi-stakeholder game dynamics. Background Technology

[0002] my country's electric vehicle industry has experienced explosive growth, forming a new type of power-transportation coupled network with coordinated development of vehicles and charging infrastructure. While the large-scale integration of electric vehicles improves energy efficiency, the randomness of their spatiotemporal distribution and the strong coupling of their charging and discharging behaviors lead to new challenges for the distribution network, such as increased uncertainty on both the source and load sides and intensified conflicts of interest among multiple stakeholders. Especially in urban core areas, the combined effect of traffic congestion and surging charging demand has significantly altered the traditional load curve, making it imperative for distribution network planning to break through the technical paradigms of static modeling and single-stakeholder optimization.

[0003] The inventors discovered that, under the aforementioned technological background, existing power distribution network planning technologies are ill-suited to the development needs of new power systems: while time-of-use pricing mechanisms guide charging behavior through time-based segmentation, they lack real-time responses to dynamic traffic congestion and vehicle owners' charging intentions, resulting in load forecasting deviations generally exceeding 15%; centralized optimization models fail to consider the game strategies of multiple stakeholders such as power generators, operators, and users, which can easily lead to imbalances in the distribution of benefits in V2G bidirectional charging and discharging scenarios; existing planning systems mostly employ deterministic scenario analysis, making it difficult to quantify uncertainties such as fluctuations in renewable energy output and the spatiotemporal migration of charging demand, resulting in both insufficient equipment utilization and redundancy. Summary of the Invention

[0004] To overcome the shortcomings of the prior art, this invention provides a method and system for planning the structure of a power distribution network that considers multi-party game theory. This method accurately simulates the game behavior of multiple parties in the market, ensuring that each market participant continuously optimizes its own decisions during the game process to maximize its own benefits, thereby enhancing market vitality and the effectiveness and rationality of planning decisions.

[0005] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: The first aspect of this invention provides a method for planning the structure and form of a power distribution network that takes into account multi-agent game theory.

[0006] Distribution network structure planning methods considering multi-agent game theory include: The target groups include power distribution network operators, charging station operators, and EV owners. To maximize the profits of charging station operators, a charging station operator model is built based on the electricity sales revenue, electricity purchase cost, operation and maintenance cost, and charging congestion penalty. To minimize the cost for distribution network operators, a distribution network operator model is constructed. To minimize charging costs and time delay costs, an objective function for EV owner decision-making is constructed. By treating EV owners as subordinate participants in the game and power grid operators and charging station operators as the main players, we can solve for the equilibrium state of the game. Based on the game equilibrium state, the lines and structures to be built in the distribution network are determined.

[0007] The second aspect of the present invention provides a power distribution network structure planning system that takes into account multi-agent game theory.

[0008] Distribution network structure planning systems considering multi-agent game dynamics include: The multi-subject determination module is configured to: determine multiple subject objects including power distribution network operators, charging station operators, and EV owners; The model building module is configured to: consider maximizing the revenue of charging station operators, build a charging station operator model based on charging station electricity sales revenue, electricity purchase cost, operation and maintenance cost, and charging congestion penalty; consider minimizing the cost of distribution network operators, build a distribution network operator model; and consider minimizing charging cost and time delay cost, build an EV owner decision objective function. The solution module is configured to: treat EV owners as subordinate participants in the game, and the decisions of power grid operators and charging station operators as the main players in the game, and solve for the equilibrium state of the game. The module for determining the lines to be built is configured to: determine the lines and structures to be built in the distribution network based on the game equilibrium state. A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in the distribution network structure planning method considering multi-agent game as described in the first aspect of the present invention.

[0009] The fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the distribution network structure planning method considering multi-agent game as described in the first aspect of the present invention.

[0010] The above one or more technical solutions have the following beneficial effects: This invention constructs a planning and decision-making model for three different stakeholders: charging station operators, power distribution network operators, and EV owners. Based on the transmission relationship among the three stakeholders, it analyzes the game behavior among them and solves the iterative process between the strategies of power distribution network operators, charging station operators, and EV owners to obtain the optimal power distribution network structure. At the same time, through the continuous optimization of their own decisions during the game process, charging station operators and EV owners maximize their respective benefits.

[0011] To address the problem that traditional time-of-use pricing mechanisms and static load forecasting models struggle to perceive real-time changes in traffic congestion and shifts in vehicle owner charging preferences, leading to excessively high prediction errors in the spatiotemporal distribution of charging load and increased risks of distribution network capacity mismatch and node voltage exceedances, this invention establishes a charging station operator model based on charging station electricity sales revenue, electricity purchase cost, operation and maintenance cost, and charging congestion penalty. This model minimizes distribution network operator costs and, consequently, charging costs and time delay costs. An objective function for EV owner decisions is also established. Game theory is incorporated into the planning of incremental distribution networks to accurately simulate the game behavior of market participants, ensuring that each participant continuously optimizes their decisions to maximize their own gains and enhancing market vitality and the effectiveness of planning decisions.

[0012] In V2G two-way interactive scenarios, power generators pursue maximizing consumption revenue, operators focus on improving equipment utilization, and users prioritize optimal charging costs, creating a strategic game. Existing centralized optimization models neglect the non-cooperative game characteristics of multiple stakeholders, leading to imbalances in benefit distribution and decreased equipment coordination efficiency in actual implementation. This invention, based on multi-stakeholder game theory, improves the rationality of planning decisions through proactive optimization within a game theory-based planning model. When analyzing the static game behavior of multiple stakeholders in power distribution network planning, EV car owners receive electricity price signals and traffic congestion data, and generate charging decisions with the goal of minimizing charging costs plus time delay costs. An improved Nesterov accelerated gradient search iterative method is used, combined with maximum and minimum points, to solve for the Nash equilibrium point of the game. Based on the game equilibrium state, the lines to be built in the power distribution network are determined, and then the structure of the power distribution network is determined.

[0013] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0014] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0015] Figure 1 This is a flowchart of the method in Example 1.

[0016] Figure 2 This is a diagram illustrating the decision-making transmission relationships among multiple subjects in Example 1.

[0017] Figure 3 This is a diagram illustrating the iterative process of multi-subject object decision-making in Example 1.

[0018] Figure 4 This is a flowchart for solving the game equilibrium state in Example 1. Detailed Implementation

[0019] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0020] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0021] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0022] The overall concept proposed in this invention is as follows: This invention incorporates game theory concepts into the planning of incremental distribution networks, proposing a method for planning the structural form of incremental distribution networks that considers multi-player game dynamics. By constructing and simulating the interactions of multiple market players and analyzing their game behavior, each market player continuously adjusts and optimizes their decision-making strategies during the game process, thereby maximizing their own interests.

[0023] Specifically, a charging station operator model with a congestion penalty mechanism and a distribution network operator cost minimization model are constructed to minimize the cost of distribution network operators and maximize the revenue of charging station operators. Analyze the decision transmission relationships among multiple subjects, and based on these relationships, determine the multiple iteration processes among the distribution network operator strategy, charging station operator strategy, and EV owner strategy. This paper analyzes the static game behavior of multiple stakeholders in power distribution network planning, taking EV owners as subordinate participants and power distribution network operators and charging station operators as the main players. Then, it solves the game equilibrium state when the cost of power distribution network operators no longer decreases and the revenue of charging station operators no longer increases during multiple iterations. Based on the game equilibrium state, the lines to be built in the distribution network are determined, and then the structure of the distribution network is determined.

[0024] The method of this invention aims to achieve three objectives: first, to minimize the investment cost of incremental distribution network planning; second, to minimize the operating costs of charging station operators; and finally, to reduce the charging costs of electric vehicle clusters as much as possible. Through this multi-stakeholder game-theoretic optimization mechanism, not only can market vitality be effectively enhanced, but the rationality and enforceability of planning decisions can also be strengthened, ultimately promoting the efficient and coordinated development of distribution network planning and electric vehicle charging infrastructure.

[0025] Example 1 This embodiment discloses a distribution network structure planning method that considers multi-agent game theory.

[0026] like Figure 1 As shown, the distribution network structure planning method considering multi-agent game theory includes: The target groups include power distribution network operators, charging station operators, and EV owners. To maximize the profits of charging station operators, a charging station operator model is built based on the electricity sales revenue, electricity purchase cost, operation and maintenance cost, and charging congestion penalty. To minimize the cost for distribution network operators, a distribution network operator model is constructed. To minimize charging costs and time delay costs, an objective function for EV owner decision-making is constructed. By treating EV owners as subordinate participants in the game and power grid operators and charging station operators as the main players, we can solve for the equilibrium state of the game. Based on the game equilibrium state, the lines and structures to be built in the distribution network are determined.

[0027] The core of the static game among electric vehicles, charging stations, and the power distribution network lies in the mutually constraining equilibrium relationship among them based on their respective economic interests and technological constraints in decisions such as charging pricing and load allocation. EV owners, aiming to minimize charging costs, tend to choose charging times with low electricity prices, but are limited by travel demand; charging station operators need to balance service fee revenue with operating costs, attracting users through pricing strategies while also considering expenses such as equipment maintenance and grid interaction; power distribution network operators need to maintain stable grid operation and guide the distribution of charging load through time-of-use pricing.

[0028] This invention integrates game theory and robust optimization principles into the planning of incremental distribution networks, proposing a distribution network structure planning method that considers multi-stakeholder game dynamics. This method constructs a planning decision-making model involving charging station operators, distribution network operators, and EV owners, representing different stakeholders. Then, it analyzes the game behavior among these three stakeholders based on their inter-stakeholder relationships. Each party adopts different strategies to pursue its own interests, engaging in a game of strategy. This embodiment improves the effectiveness of the three-party planning by analyzing the multi-stakeholder static game behavior in incremental distribution network planning.

[0029] The method described in this embodiment will now be explained in detail.

[0030] In this embodiment, the power distribution network structure planning method considering multi-agent game theory specifically includes: 1) Collect accurate electric vehicle operation data and charging / discharging data models to construct accurate electric vehicle charging / discharging models; 2) Construct charging station operator models and power distribution network operator models; 3) Analyze the multi-agent static game behavior in incremental distribution network planning; 4) Construct a planning model solution process based on game theory.

[0031] 1. Electric vehicle model Electric vehicles are powered by batteries, and can be viewed as mobile energy storage devices that move energy according to the driver's driving characteristics.

[0032] 1.1 Objective Function (1) in , The electricity prices for EV charging and V2G discharging are respectively. , For charging and discharging power, The battery loss function is calculated using the power throughput method. TrafficDelay(t) is the time value coefficient, representing the delay time caused by traffic congestion, and is dynamically updated based on the real-time traffic API.

[0033] 1.2 Constraints The remaining charge status of a battery is usually expressed as its state of charge (SOC). (Time period) The initial electric vehicle's battery level is The SOC at the start of the next time period is: (2) In the formula: , For charging and discharging efficiency, , For charging and discharging power, Let t be the charge level of the electric vehicle at time t; Let be the charge level of the electric vehicle at time t-1; This refers to the time interval.

[0034] To ensure the lifespan of the battery, the charging power and SOC of an electric vehicle must meet the following constraints: (3) (4) In the formula: , For the maximum and minimum charging power of electric vehicles, , These represent the upper and lower limits of the State of Charge (SOC) for electric vehicles.

[0035] 2. Charging station operators 2.1 Objective Function (5) (6) in, , , and These represent the revenue from electricity sales at charging stations, electricity purchase costs, operation and maintenance costs, and congestion penalty costs, respectively. Indicates the EV charging level. This indicates the electricity price for sale. This represents the construction cost per unit capacity. For 0-1 variables, Indicates the points to be selected Connect to charging station This indicates the charging station capacity at candidate point s. Indicates the points to be selected Not connected to charging stations For the discount rate, For the lifespan of the charging station. The power generation operation and maintenance cost per unit of electricity generated; This represents the number of candidate nodes s that can be connected to the charging station. Let be the queue length at station s during time period t. This is the congestion premium coefficient. The longest queue number, Fixed penalty parameters; Indicates the operating cycle.

[0036] 2.2 Constraints (7) Indicates the point to be selected The minimum capacity of the charging station. Indicates the point to be selected The maximum capacity of the charging station.

[0037] 3. Distribution network operators 3.1 Objective Function (8) (9) (8) Characterizes the cost of supplying power from the distribution network to the EV cluster. Among them, , and These represent the real-time electricity purchase cost, energy storage usage cost, and photovoltaic usage cost for the distribution network operator, respectively. This represents the charging charge level of the i-th EV. This indicates the time-of-use electricity price. and These represent the loss coefficient of the energy storage system and the unit photovoltaic power generation cost coefficient, respectively. for Energy storage charging and discharging power at any time, for Photovoltaic power generation at any given time Indicates the number of power distribution network operators; Indicates the amount of energy stored; This indicates the number of photovoltaic cells.

[0038] 3.2 Constraints The constraints include new line constraints (10), power flow constraints (11), and voltage and power constraints (12).

[0039] (10) For the set of newly added load nodes, For new load nodes Line under construction 0-1 variables.

[0040] (11) and They are respectively Time Node Active power and reactive power, and They are respectively Time Node and nodes The voltage amplitude. and They are nodes and nodes The conductivity and susceptance between them and They are nodes and nodes The sine and cosine values ​​of the voltage phase angle difference between them.

[0041] (12) and They are respectively Time Node With nodes Transmission power and nodes With nodes The maximum power transmitted between them. , and They are respectively Time Node The voltage amplitude, upper and lower voltage limits.

[0042] 4. Multi-stakeholder Game Theory in Incremental Distribution Network Planning In this embodiment, the market participants are charging station operators, power grid operators, and EV owners. The decision-making relationship among these three is as follows: Figure 2 As shown.

[0043] The assets of distribution network operators also include distributed power sources such as photovoltaics and energy storage. Due to the uncertainty of their output, the safety and reliability of distribution network operation will be reduced, and additional operating costs will be incurred, leading to a decrease in its economic efficiency.

[0044] like Figure 2 As shown, charging station operators locate and select capacity for charging stations within the current power distribution network architecture. EV owners charge their vehicles at different times and locations based on time-of-use pricing. Power distribution network operators expand the lines by receiving information from other entities, minimizing costs and voltage / power constraints, thereby forming a new power distribution network structure.

[0045] Since EV owners can only passively choose charging time and location and cannot directly influence the power distribution network structure and charging station site selection, the EV cluster makes decisions based on charging price information provided by charging stations. Therefore, the optimal solution for owners mainly focuses on how to choose charging time and charging station to minimize charging costs. Charging stations incentivize owners to choose suitable charging times through pricing, while the power distribution network optimizes the scheduling between power supply capacity and power demand to ensure that charging demand is met and avoid system overload.

[0046] Therefore, although car owners do not directly participate in the site selection and capacity decisions of the power distribution network and charging stations, their charging decisions still affect the overall optimization of the system. Considering how car owners can minimize charging prices by choosing the optimal charging station and time will help to further improve the game theory model of the power distribution network and charging stations, thereby achieving collaborative optimization among multiple stakeholders.

[0047] In the process of independent decision-making and joint promotion of incremental distribution network planning and construction, charging station operators and distribution network operators, having mastered all of each other's strategies, have no sequential order of decision-making and actions. This creates a static game relationship among EV owners, distribution network operators, and charging station operators, as illustrated in the iterative process diagram. Figure 3 Show.

[0048] EV owners dynamically optimize their charging strategies through reinforcement learning algorithms: Real-time electricity price signals and traffic congestion data are received to generate charging decisions with the goal of minimizing charging costs plus time delay costs, forming an adaptive feedback loop. The decision objective is:

[0049] in , The electricity prices for EV charging and V2G discharging are respectively. The battery loss function is calculated using the power throughput method. TrafficDelay(t) is the time value coefficient, representing the delay time caused by traffic congestion, and is dynamically updated based on the real-time traffic API.

[0050] In one iteration, EV owners autonomously plan their charging routes, generating charging loads and uploading them to the power grid operator. Based on the previous site selection and capacity allocation schemes of the charging station operators, the power grid operator optimizes line configurations to minimize grid operating costs. Meanwhile, the charging station operators optimize site selection and capacity allocation based on new grid line construction plans to improve efficiency. Both parties engage in an iterative game by dynamically adjusting network topology and charging facility parameters until the system reaches equilibrium.

[0051] During the iteration process, if the costs for both parties in the game—the distribution network operator and the charging station operator—no longer decrease and their revenues no longer increase, then the game is considered to have reached an equilibrium state. Specifically, this can be represented as: (13) and The strategies of charging station operators and distribution network operators under game equilibrium states, respectively. This represents the set of values ​​at which the objective function is minimized. This represents the function for calculating the revenue of charging station operators. This represents the function for calculating the investment costs of a power distribution network operator.

[0052] 5. Static Joint Game Process Step 1: Input the raw data.

[0053] The raw data includes distribution network structure parameters, distribution network flexibility resource parameters, EV cluster operation parameters, and charging station capacity parameters.

[0054] Distribution network structural parameters include the impedance between various nodes in the distribution network. In addition to straight-line distance, distribution network flexibility resource parameters include photovoltaic capacity and energy storage capacity.

[0055] Step 2: Generate the game strategy space of the main participants.

[0056] The game strategy space of the main participants specifically refers to the capacity configuration of charging station operators at each node, as well as the structural parameters of the power distribution network.

[0057] Step 3: Obtain initial values , and .

[0058] Randomly select a set of values ​​from the strategy spaces of charging station operators and distribution network operators respectively. , Used as the initial value for iteration. To meet the charging needs of electric vehicle clusters at various nodes.

[0059] Step 4: Each participating entity performs independent optimization.

[0060] Distribution network operators iterate their strategies based on the previous site selection and capacity allocation schemes of charging station operators, aiming to minimize grid operating costs; while charging station operators optimize site selection and capacity allocation based on the new grid line construction schemes, aiming to maximize electricity sales revenue.

[0061] Taking the nth round of the game as an example, the charging station investment operator bases its strategy on the distribution network investment operator's strategy after n-1 rounds of the game. Based on the distribution network architecture, the optimal location and capacity determination strategy for charging stations is obtained with the goal of maximizing economic benefits. ; EV owners can determine their charging strategies based on the current location of charging stations, aiming to minimize charging costs plus time delay costs. The EV charging power is summed up to obtain the required power deficit of the distribution network, and the distribution network investment and operation strategies are adjusted according to the power deficit. .

[0062] Step 5: Determine whether the game has reached an equilibrium state.

[0063] An improved Nesterov accelerated gradient method is used to solve for the Nash equilibrium. The iterative formula is as follows: (14) Where n represents the nth round, n-1 represents the (n-1)th round, and n-2 represents the (n-2)th round; f represents the optimal site selection and capacity allocation strategy for charging stations, and y represents the strategy of distribution network investment operators; This represents the optimal location and capacity allocation strategy for the charging station in the (n-1)th round; This represents the optimal location and capacity allocation strategy for the charging station in the (n-2)th round; This represents the distribution network investment operator strategy in the (n-1)th round; This represents the distribution network investment operator strategy in the (n-2)th round; This represents the suboptimal solution of the charging station location and capacity gradation strategy obtained in the nth round using gradient optimization; This represents the suboptimal solution of the distribution network investment operator strategy obtained in the nth round using gradient optimization; This represents the momentum coefficient for the nth round.

[0064] Set initial values ​​for momentum coefficient As the number of game rounds increases, According to the formula, it decreases monotonically, where... for The set empirical parameter can be set to 0, when and At that time, it is believed that the game has reached equilibrium. It represents a very small positive value.

[0065] Step 6: Output the final equilibrium solution .

[0066] Step 7: Compare and analyze the necessity of multi-agent game theory.

[0067] A comparative analysis is conducted using two different power distribution network planning schemes: one incorporating multi-stakeholder game theory and the other not. The comparison focuses on three key differences: first, the investment costs of power distribution network operators; second, the profitability of charging station operators; and third, the charging costs of electric vehicle clusters. This demonstrates the necessity of considering multi-stakeholder game theory in the method presented in this embodiment.

[0068] After obtaining the final equilibrium solution, based on the equilibrium solution of the charging station operator... This determines the charging station operator's decision-making, namely the installation location, quantity, and capacity of charging piles; based on the equilibrium solution of the distribution network operator. This determines the distribution network operator's decision, namely, the new route for the distribution network. The new route refers to expanding the distribution network based on the game theory outcome, connecting previously unconnected nodes, and obtaining parameters including the nodes to be expanded and the straight-line distance between any two nodes.

[0069] Example 2 This embodiment discloses a power distribution network structure planning system that considers multi-agent game theory.

[0070] Distribution network structure planning systems considering multi-agent game dynamics include: The multi-subject determination module is configured to: determine multiple subject objects including power distribution network operators, charging station operators, and EV owners; The model building module is configured to: consider maximizing the revenue of charging station operators, build a charging station operator model based on charging station electricity sales revenue, electricity purchase cost, operation and maintenance cost, and charging congestion penalty; consider minimizing the cost of distribution network operators, build a distribution network operator model; and consider minimizing charging cost and time delay cost, build an EV owner decision objective function. The solution module is configured to: treat EV owners as subordinate participants in the game, and the decisions of power grid operators and charging station operators as the main players in the game, and solve for the equilibrium state of the game. The module for determining the lines to be built is configured to: determine the lines and structures to be built in the distribution network based on the game equilibrium state.

[0071] The model building module in this embodiment: 1. Electric vehicle model Electric vehicles are powered by batteries, and can be viewed as mobile energy storage devices that move energy according to the driver's driving characteristics.

[0072] 1.1 Objective Function (1) in , The electricity prices for EV charging and V2G discharging are respectively. , For charging and discharging power, The battery loss function is calculated using the power throughput method. TrafficDelay(t) is the time value coefficient, representing the delay time caused by traffic congestion, and is dynamically updated based on the real-time traffic API.

[0073] 1.2 Constraints The remaining charge status of a battery is usually expressed as its state of charge (SOC). (Time period) The initial electric vehicle's battery level is The SOC at the start of the next time period is: (2) In the formula: , For charging and discharging efficiency, , For charging and discharging power, Let t be the charge level of the electric vehicle at time t; Let be the charge level of the electric vehicle at time t-1; This refers to the time interval.

[0074] To ensure the lifespan of the battery, the charging power and SOC of an electric vehicle must meet the following constraints: (3) (4) In the formula: , For the maximum and minimum charging power of electric vehicles, , These represent the upper and lower limits of the State of Charge (SOC) for electric vehicles.

[0075] 2. Charging station operators 2.1 Objective Function (5) (6) in, , , and These represent the revenue from electricity sales at charging stations, electricity purchase costs, operation and maintenance costs, and congestion penalty costs, respectively. Indicates the EV charging level. This indicates the electricity price for sale. This represents the construction cost per unit capacity. For 0-1 variables, Indicates the points to be selected Connect to charging station This indicates the charging station capacity at candidate point s. Indicates the points to be selected Not connected to charging stations For the discount rate, For the lifespan of the charging station. The power generation operation and maintenance cost per unit of electricity generated; This represents the number of candidate nodes s that can be connected to the charging station. Let be the queue length at station s during time period t. This is the congestion premium coefficient. The longest queue number, Fixed penalty parameters; Indicates the operating cycle.

[0076] 2.2 Constraints (7) Indicates the point to be selected The minimum capacity of the charging station. Indicates the point to be selected The maximum capacity of the charging station.

[0077] 3. Distribution network operators 3.1 Objective Function (8) (9) (8) Characterizes the cost of supplying power from the distribution network to the EV cluster. Among them, , and These represent the real-time electricity purchase cost, energy storage usage cost, and photovoltaic usage cost for the distribution network operator, respectively. This represents the charging charge level of the i-th EV. This indicates the time-of-use electricity price. and These represent the loss coefficient of the energy storage system and the unit photovoltaic power generation cost coefficient, respectively. for Energy storage charging and discharging power at any time, for Photovoltaic power generation at any given time Indicates the number of power distribution network operators; Indicates the amount of energy stored; This indicates the number of photovoltaic cells.

[0078] 3.2 Constraints The constraints include new line constraints (10), power flow constraints (11), and voltage and power constraints (12).

[0079] (10) For the set of newly added load nodes, For new load nodes Line under construction 0-1 variables.

[0080] (11) and They are respectively Time Node Active power and reactive power, and They are respectively Time Node and nodes The voltage amplitude. and They are nodes and nodes The conductivity and susceptance between them and They are nodes and nodes The sine and cosine values ​​of the voltage phase angle difference between them.

[0081] (12) and They are respectively Time Node With nodes Transmission power and nodes With nodes The maximum power transmitted between them. , and They are respectively Time Node The voltage amplitude, upper and lower voltage limits.

[0082] 4. Multi-stakeholder Game Theory in Incremental Distribution Network Planning In this embodiment, the market participants are charging station operators, power grid operators, and EV owners. The decision-making relationship among these three is as follows: Figure 2 As shown.

[0083] The assets of distribution network operators also include distributed power sources such as photovoltaics and energy storage. Due to the uncertainty of their output, the safety and reliability of distribution network operation will be reduced, and additional operating costs will be incurred, leading to a decrease in its economic efficiency.

[0084] like Figure 2 As shown, charging station operators locate and select capacity for charging stations within the current power distribution network architecture. EV owners charge their vehicles at different times and locations based on time-of-use pricing. Power distribution network operators expand the lines by receiving information from other entities, minimizing costs and voltage / power constraints, thereby forming a new power distribution network structure.

[0085] Since EV owners can only passively choose charging time and location and cannot directly influence the power distribution network structure and charging station site selection, the EV cluster makes decisions based on charging price information provided by charging stations. Therefore, the optimal solution for owners mainly focuses on how to choose charging time and charging station to minimize charging costs. Charging stations incentivize owners to choose suitable charging times through pricing, while the power distribution network optimizes the scheduling between power supply capacity and power demand to ensure that charging demand is met and avoid system overload.

[0086] Therefore, although car owners do not directly participate in the site selection and capacity decisions of the power distribution network and charging stations, their charging decisions still affect the overall optimization of the system. Considering how car owners can minimize charging prices by choosing the optimal charging station and time will help to further improve the game theory model of the power distribution network and charging stations, thereby achieving collaborative optimization among multiple stakeholders.

[0087] In the process of independent decision-making and joint promotion of incremental distribution network planning and construction, charging station operators and distribution network operators, having mastered all of each other's strategies, have no sequential order of decision-making and actions. This creates a static game relationship among EV owners, distribution network operators, and charging station operators, as illustrated in the iterative process diagram. Figure 3 Show.

[0088] EV owners dynamically optimize their charging strategies through reinforcement learning algorithms: Real-time electricity price signals and traffic congestion data are received to generate charging decisions with the goal of minimizing charging costs plus time delay costs, forming an adaptive feedback loop. The decision objective is:

[0089] in , The electricity prices for EV charging and V2G discharging are respectively. The battery loss function is calculated using the power throughput method. TrafficDelay(t) is the time value coefficient, representing the delay time caused by traffic congestion, and is dynamically updated based on the real-time traffic API.

[0090] In one iteration, EV owners autonomously plan their charging routes, generating charging loads and uploading them to the power grid operator. Based on the previous site selection and capacity allocation schemes of the charging station operators, the power grid operator optimizes line configurations to minimize grid operating costs. Meanwhile, the charging station operators optimize site selection and capacity allocation based on new grid line construction plans to improve efficiency. Both parties engage in an iterative game by dynamically adjusting network topology and charging facility parameters until the system reaches equilibrium.

[0091] During the iteration process, if the costs for both parties in the game—the distribution network operator and the charging station operator—no longer decrease and their revenues no longer increase, then the game is considered to have reached an equilibrium state. Specifically, this can be represented as: (13) and The strategies of charging station operators and distribution network operators under game equilibrium states, respectively. This represents the set of values ​​at which the objective function is minimized. This function represents the calculation function for the operating revenue of charging station operators. This represents the function for calculating the investment costs of a power distribution network operator.

[0092] 5. Static Joint Game Process Step 1: Input the raw data.

[0093] The raw data includes distribution network structure parameters, distribution network flexibility resource parameters, EV cluster operation parameters, and charging station capacity parameters.

[0094] Distribution network structural parameters include the impedance between various nodes in the distribution network. In addition to straight-line distance, distribution network flexibility resource parameters include photovoltaic capacity and energy storage capacity.

[0095] Step 2: Generate the game strategy space of the main participants.

[0096] The game strategy space of the main participants specifically refers to the capacity configuration of charging station operators at each node, as well as the structural parameters of the power distribution network.

[0097] Step 3: Obtain initial values , and .

[0098] Randomly select a set of values ​​from the strategy spaces of charging station operators and distribution network operators respectively. , Used as the initial value for iteration. To meet the charging needs of electric vehicle clusters at various nodes.

[0099] Step 4: Each participating entity performs independent optimization.

[0100] Distribution network operators iterate their strategies based on the previous site selection and capacity allocation schemes of charging station operators, aiming to minimize grid operating costs; while charging station operators optimize site selection and capacity allocation based on the new grid line construction schemes, aiming to maximize electricity sales revenue.

[0101] Taking the nth round of the game as an example, the charging station investment operator bases its strategy on the distribution network investment operator's strategy after n-1 rounds of the game. Based on the distribution network architecture, the optimal location and capacity determination strategy for charging stations is obtained with the goal of maximizing economic benefits. ; EV owners can determine their charging strategies based on the current location of charging stations, aiming to minimize charging costs plus time delay costs. The EV charging power is summed up to obtain the required power deficit of the distribution network, and the distribution network investment and operation strategies are adjusted according to the power deficit. .

[0102] Step 5: Determine whether the game has reached an equilibrium state.

[0103] An improved Nesterov accelerated gradient method is used to solve for the Nash equilibrium. The iterative formula is as follows: (14) Where n represents the nth round, n-1 represents the (n-1)th round, and n-2 represents the (n-2)th round; f represents the optimal site selection and capacity allocation strategy for charging stations, and y represents the strategy of distribution network investment operators; This represents the optimal location and capacity allocation strategy for the charging station in the (n-1)th round; This represents the optimal location and capacity allocation strategy for the charging station in the (n-2)th round; This represents the distribution network investment operator strategy in the (n-1)th round; This represents the distribution network investment operator strategy in the (n-2)th round; This represents the suboptimal solution of the charging station location and capacity gradation strategy obtained in the nth round using gradient optimization; This represents the suboptimal solution of the distribution network investment operator strategy obtained in the nth round using gradient optimization; This represents the momentum coefficient for the nth round.

[0104] Set initial values ​​for momentum coefficient As the number of game rounds increases, According to the formula, it decreases monotonically, where... for The set empirical parameter can be set to 0, when and At that time, it is believed that the game has reached equilibrium. It represents a very small positive value.

[0105] Step 6: Output the final equilibrium solution .

[0106] Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0107] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the distribution network structure planning method considering multi-agent game as described in Embodiment 1 of this disclosure.

[0108] Example 4 The purpose of this embodiment is to provide an electronic device.

[0109] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the distribution network structure planning method considering multi-agent game as described in Embodiment 1 of this disclosure.

[0110] The steps and methods involved in the apparatuses of Embodiments 2, 3, and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0111] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0112] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A distribution network structure planning method considering multi-agent game theory, characterized in that, include: The target groups include power distribution network operators, charging station operators, and EV owners. To maximize the profits of charging station operators, a charging station operator model is built based on the electricity sales revenue, electricity purchase cost, operation and maintenance cost, and charging congestion penalty. To minimize the cost for distribution network operators, a distribution network operator model is constructed. To minimize charging costs and time delay costs, an objective function for EV owner decision-making is constructed. By treating EV owners as subordinate participants in the game and power grid operators and charging station operators as the main players, we can solve for the equilibrium state of the game. Based on the game equilibrium state, the lines and structures to be built in the distribution network are determined.

2. The distribution network structure planning method considering multi-agent game theory as described in claim 1, characterized in that, The objective function for EV owner decision-making is: ； In the formula: , The electricity prices are for EV charging and V2G discharging, respectively. , For charging and discharging power, The battery loss function is calculated using the power throughput method. TrafficDelay(t) is the time value coefficient, representing the delay time caused by traffic congestion, and is dynamically updated based on the real-time traffic API.

3. The distribution network structure planning method considering multi-agent game theory as described in claim 1, characterized in that, The charging station operator model is as follows: ； ； in, , , and These represent the revenue from electricity sales at charging stations, electricity purchase costs, operation and maintenance costs, and congestion penalty costs, respectively. Indicates the EV charging capacity. Indicates the electricity price; This represents the construction cost per unit capacity. For 0-1 variables, Indicates the points to be selected Connect to charging station Indicates the points to be selected Not connected to charging stations This indicates the charging station capacity at candidate point s. For the discount rate, For the lifespan of the charging station. The power generation operation and maintenance cost per unit of electricity generated; This represents the number of candidate nodes s that can be connected to the charging station. Let be the queue length at station s during time period t. This is the congestion premium coefficient. The longest queue number, Fixed penalty parameters; Indicates the operating cycle.

4. The distribution network structure planning method considering multi-agent game theory as described in claim 1, characterized in that, The distribution network operator model is as follows: ； ； in, , and These represent the electricity sales revenue of the distribution network operator, the cost of energy storage, and the cost of photovoltaic use, respectively. This represents the charging power of the i-th EV. Indicates time-of-use electricity pricing; and These represent the loss coefficient of the energy storage system and the unit photovoltaic power generation cost coefficient, respectively. for Energy storage charging and discharging power at any time for Photovoltaic power generation at any given time; Indicates the number of power distribution network operators; Indicates the amount of energy stored; This indicates the number of photovoltaic cells.

5. The distribution network structure planning method considering multi-agent game theory as described in claim 1, characterized in that, This also includes determining the constraints for the EV owner's decision objective function, the charging station operator model, and the power distribution network operator model, among which: The constraints of the objective function for EV owner decision-making include: ； ； ； In the formula: , For charging and discharging efficiency, , For charging and discharging power, This refers to the average capacity of electric vehicle batteries. Let t be the charge level of the electric vehicle at time t; Let be the charge level of the electric vehicle at time t-1; For time intervals; , Maximum and minimum charging power for electric vehicles; , These are the upper and lower limits of the State of Charge (SOC) for electric vehicles. The constraints of the charging station operator model include: ； Indicates the point to be selected The minimum capacity of the charging station, Indicates the point to be selected The maximum capacity of the charging station; Indicates the EV charging level; The constraints of the distribution network operator model include new line constraints, power flow constraints, and voltage and power constraints, which are as follows: ； For the set of newly added load nodes, For new load nodes Line under construction 0-1 variables; ； and They are respectively Time Node Active power and reactive power; and They are respectively Time Node and nodes The voltage amplitude; and They are nodes and nodes The conductivity and susceptance between them and They are nodes and nodes Voltage phase angle difference between The sine and cosine values; ； and They are respectively Time Node With nodes Transmission power between; , and They are respectively Time Node The voltage amplitude, upper voltage limit, and lower voltage limit.

6. The distribution network structure planning method considering multi-agent game theory as described in claim 1, characterized in that, Also includes: Based on the decision-making transmission relationship among multiple subjects, the multiple iteration process between the power distribution network operator strategy, the charging station operator strategy, and the EV owner strategy is determined. Solve the game equilibrium state when the cost of the distribution network operator no longer decreases and the revenue of the charging station operator no longer increases during multiple iterations.

7. The distribution network structure planning method considering multi-agent game theory as described in claim 6, characterized in that, The decision-making transmission relationship among the multiple subjects is specifically as follows: Under the current power grid structure, charging station operators should select and determine the location and number of charging piles to maximize their profits. EV owners can choose the time and place to charge based on the installation location of the charging piles and the time-of-use electricity price determined by the current charging station operator, so as to minimize charging costs and time delay costs. Based on the current charging station operators' site selection and capacity allocation plans, distribution network operators expand distribution network lines to minimize operating costs, forming the next round of distribution network structure.

8. The distribution network structure planning method considering multi-agent game theory as described in claim 7, characterized in that, This also includes EV owners dynamically optimizing charging strategies through reinforcement learning algorithms, specifically: EV owners receive real-time electricity price signals and traffic congestion data to generate charging decisions with the goal of minimizing charging costs and time delays, forming an adaptive feedback loop.

9. The distribution network structure planning method considering multi-agent game theory as described in claim 1, characterized in that, The game process is as follows: Obtain raw parameters, including distribution network structure parameters, distribution network flexibility resource parameters, EV cluster operation parameters, and charging station capacity parameters; Construct a set of nodes for charging stations to be built and a set of power distribution lines to be built, and generate a game strategy space for the main participants. A set of values ​​is randomly selected from the strategy space of the main participants as the initial values ​​for iteration; Independent optimization for each participating entity: The distribution network operator iterates its strategy based on the previous site selection and capacity determination scheme of the charging station operator, with the goal of minimizing the grid operating cost; Based on the proposed new power grid lines, charging station operators optimize site selection and capacity configuration with the goal of maximizing electricity sales revenue. Electric vehicle owners, based on the game between charging station operators and charging station operators, determine the charging strategy that minimizes charging costs and time delay costs. Determine whether the game has reached an equilibrium state. If it has not reached an equilibrium state, continue iterating until the game reaches an equilibrium state. Based on the game equilibrium state of the charging station operator strategy, the optimal installation location and number of charging piles are determined, and then the power distribution network structure is determined.

10. A power distribution network structure planning system considering multi-agent game theory, characterized in that, include: The multi-subject determination module is configured to: determine multiple subject objects including power distribution network operators, charging station operators, and EV owners; The model building module is configured to: consider maximizing the revenue of charging station operators, build a charging station operator model based on charging station electricity sales revenue, electricity purchase cost, operation and maintenance cost, and charging congestion penalty; consider minimizing the cost of distribution network operators, build a distribution network operator model; and consider minimizing charging cost and time delay cost, build an EV owner decision objective function. The solution module is configured to: treat EV owners as subordinate participants in the game, and the decisions of power grid operators and charging station operators as the main players in the game, and solve for the equilibrium state of the game. The module for determining the lines to be built is configured to: determine the lines and structures to be built in the distribution network based on the game equilibrium state.

11. A computer-readable storage medium having a program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the distribution network structure planning method considering multi-agent game as described in any one of claims 1-9.

12. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the distribution network structure planning method considering multi-agent game as described in any one of claims 1-9.

Images