Collaborative planning method for electric vehicle charging infrastructure and urban power distribution network

Abstract

The invention relates to the technical field of power grid collaboration, in particular to an electric vehicle charging infrastructure and urban power distribution network collaborative planning method, which comprises the following steps of: acquiring historical electric vehicle charging demand data of an existing charging pile and urban power distribution network load data; predicting a future electric vehicle charging demand by using a machine learning algorithm to generate a prediction result; analyzing the bearing capacity of each node of the power distribution network based on the power distribution network load data and the electric vehicle charging demand prediction result, and identifying overload nodes and time periods; preliminarily determining a newly-added charging pile layout according to a bearing capacity analysis result of the power distribution network; and an optimization algorithm is adopted to solve the optimal scheme, and an optimal newly-added charging pile layout scheme is obtained. Through reasonable charging pile layout, charging infrastructure resources are utilized to the maximum extent, resource waste is avoided, the utilization efficiency of the charging piles is improved, the planning cost is effectively controlled, and balance between economical efficiency and feasibility is achieved.

Description

Technical Field

[0001] The present invention relates to the field of power grid coordination technology, and in particular to a method for coordinated planning of electric vehicle charging infrastructure and urban distribution networks. Background Art

[0002] With the popularity of electric vehicles, the demand for electric vehicle charging infrastructure has grown rapidly. However, the construction of charging piles and the load balancing of urban distribution networks have become increasingly prominent. Without effective planning and coordination, the concentrated peak of electric vehicle charging load may cause overload of the distribution network, affecting the stability and reliability of the urban power system. Traditional charging pile layout methods often ignore the dynamic balance management of urban distribution network loads, which can easily lead to local grid overload or waste of resources. The existing electric vehicle charging infrastructure planning methods mainly have the following problems:

[0003] The lack of effective integration and analysis of historical electric vehicle charging demand data and urban distribution network load data leads to inaccurate basic data for planning schemes, affecting the reliability of decision-making.

[0004] During the load assessment process, the impact of dynamic factors such as seasonality and special events was not fully considered, resulting in inaccurate assessment of the load at each node in the distribution network and inability to effectively predict and avoid overload of the distribution network.

[0005] Traditional methods often only focus on a single factor and fail to comprehensively consider the relationship between planning cost, load balance and charging demand. The lack of an overall optimization strategy makes it difficult for planning schemes to achieve optimal results in practical applications. Summary of the invention

[0006] Based on the above objectives, the present invention provides a method for collaborative planning of electric vehicle charging infrastructure and urban distribution network.

[0007] The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network includes the following steps:

[0008] S1, data collection: collect historical electric vehicle charging demand data of existing charging piles and urban distribution network load data, including charging pile usage frequency, charging time, charging power, and load conditions of each node in the distribution network. The data collection also includes data preprocessing, which includes normalizing the collected data and completing missing data.

[0009] S2, demand forecasting: Based on the preprocessed data, the machine learning algorithm is used to predict the future charging demand of electric vehicles and generate forecast results covering the charging demand and distribution in different time periods;

[0010] S3, distribution network carrying capacity analysis: Based on the distribution network load data and electric vehicle charging demand forecast results, analyze the carrying capacity of each node in the distribution network and identify overloaded nodes and time periods;

[0011] S4, planning of new charging infrastructure: based on the results of the distribution network carrying capacity analysis, preliminarily determine the layout of new charging piles, including the optimal layout location and quantity, giving priority to areas with low distribution network load and high demand for electric vehicle charging;

[0012] S5, Collaborative Optimization: Establish a collaborative optimization model for charging infrastructure planning and distribution network load balancing, comprehensively consider charging demand, distribution network load and planning cost, use optimization algorithm to solve the optimal solution, and obtain the optimal layout plan for new charging piles.

[0013] Optionally, the collection of historical electric vehicle charging demand data includes:

[0014] Charging pile usage frequency: By collecting statistics on the usage records of each charging pile, the number of times each charging pile is used in a specific time period can be obtained;

[0015] Charging time: record the start and end time of each charge, calculate the charging duration, and summarize the duration data;

[0016] Charging power: monitor and record the charging power changes during each charging process, and summarize the average charging power and peak power of each charging pile in different time periods;

[0017] The collection of load data of the urban distribution network includes:

[0018] Load conditions at each node of the distribution network: The distribution network monitoring system collects load data of each node of the distribution network in real time, including voltage, current, and power factor, and records load changes in different time periods.

[0019] Optionally, the machine learning algorithm in S2 adopts a decision tree regression model, and S2 specifically includes:

[0020] S21, extracting features from the preprocessed data, including time features (hours, weeks, months), spatial features (charging pile locations), and historical charging demand features (average charging demand in the past week), and encoding the features;

[0021] S22, dividing the preprocessed data set into a training set and a test set;

[0022] S23, select the decision tree regression model for demand forecasting. The decision tree regression model divides the data set into several subsets, builds a tree structure in each subset, and finds the best split point so that the data on each leaf node has the lowest mean square error (MSE). The goal of decision tree regression is to minimize the mean square error, which is expressed as:

[0023] Among them, y i is the actual value, is the predicted value, n is the number of samples;

[0024] S24, use the test set to evaluate the trained decision tree regression model, and use the evaluation index to evaluate the prediction performance of the model;

[0025] S25, based on the evaluation results, the model is optimized, including adjusting the depth of the decision tree, the minimum number of sample splits, and the minimum number of sample leaf nodes to prevent overfitting or underfitting and improve the prediction accuracy of the model.

[0026] Optionally, the decision tree regression model specifically includes:

[0027] Select the split point: At each node, select a feature and its corresponding split point to divide the data set into two subsets so that the samples in the split subsets are more similar. The criterion for selecting the split point is to minimize the mean square error after the split. The standard formula for setting the split point is: Among them, n L and n R are the number of samples of the left child node and the right child node respectively, and are the average values ​​of the left and right child nodes, y i and j are the actual values ​​of the left child node and the right child node respectively;

[0028] Recursive splitting: Repeatedly select split points for each subset until a preset stopping condition is reached.

[0029] Optionally, the distribution network carrying capacity analysis in S3 specifically includes:

[0030] S31, integrating the collected distribution network load data with the electric vehicle charging demand forecast results to form a comprehensive database, which includes historical load data of each node in the distribution network and predicted future charging demand data;

[0031] S32, node load dynamic adjustment: Based on historical load data and predicted charging demand data, dynamic load adjustment coefficients are introduced, including basic load dynamic adjustment coefficient α t , Dynamic adjustment coefficient of electric vehicle charging load βt , to reflect the impact of distribution network on load in different seasons and special events (holidays, large-scale activities);

[0032] S33, dynamic analysis of node carrying capacity: analyze the dynamic adjustment load of each node, consider the volatility and extreme conditions of the node load, obtain the load fluctuation range and frequency of each node through historical data analysis, and introduce the fluctuation coefficient γ to reflect the impact of node load fluctuation;

[0033] S34, Dynamic identification of overloaded nodes and time periods: Dynamic analysis of the adjusted node load, combined with the fluctuation coefficient γ i and adjust the load to determine whether the node is at risk of overload in certain periods of time. The dynamic overload determination formula is:

[0034]

[0035] in;

[0036] Adjusted_Load i,t represents the adjusted total load of node i in time period t;

[0037] Dynamic_Capacity i,t represents the dynamic carrying capacity of node i in time period t;

[0038] Overload i,t Indicates the result of determining whether node i is overloaded in time period t.

[0039] Optionally, the node load dynamic adjustment is expressed as:

[0040] Adjusted_Load i,t =Base_Load i,t ·(1+α t )+Predicted_EV_Load i,t ·

[0041] (1+β t ),in;

[0042] Adjusted_Load i,t represents the adjusted total load of node i in time period t;

[0043] Base_Load i,t represents the basic load of node i in time period t;

[0044] Predicted_EV_Load i,t represents the electric vehicle charging load predicted by node i in time period t;

[0045] α t It represents the dynamic adjustment coefficient of the base load in time period t, reflecting the impact of seasonality and special events on the base load;

[0046] β t It represents the dynamic adjustment coefficient of electric vehicle charging load in time period t, reflecting the impact of seasonality and special events on charging load.

[0047] Optionally, the introduction of the fluctuation coefficient γ to reflect the influence of node load fluctuation is expressed as:

[0048] Dynamic_Capacity i,t =Max_Capacity i ·(1-γ i ),in;

[0049] Dynamic_Capacity i,t represents the dynamic carrying capacity of node i in time period t;

[0050] Max_Capacity represents the maximum carrying capacity of node i;

[0051] γ i It represents the load fluctuation coefficient of node i, reflecting the safety margin within the node load fluctuation range.

[0052] Optionally, the S4 specifically includes:

[0053] Node priority evaluation formula:

[0054] in;

[0055] Priority i represents the priority of node i;

[0056] represents the electric vehicle charging load predicted by node i in time period t;

[0057] Dynamic Capacity i,t represents the dynamic carrying capacity of node i in time period t;

[0058] represents the adjusted total load of node i in time period t;

[0059] According to the node priority evaluation results, high-priority nodes are selected as the locations for the layout of new charging piles.

[0060] Determine the number of charging piles: Based on the charging demand forecast results and the dynamic carrying capacity of the node, determine the number of new charging piles for each priority node:

[0061] Among them, New Chargers i Indicates the number of charging piles that need to be added to node i;

[0062] Average Charger Capacity indicates the average charging capacity of a single charging pile.

[0063] Optionally, the collaborative optimization in S5 uses the preliminarily determined layout of the newly added charging piles, including the optimal layout position and quantity, as the input of the collaborative optimization model, comprehensively considering the charging demand, the distribution network load and the planning cost. The collaborative optimization model includes defining the objective function, the constraint conditions and the optimization algorithm. The optimization algorithm is used to solve the collaborative optimization model to obtain the optimal layout plan of the newly added charging piles. The optimal plan minimizes the planning cost while meeting the charging demand and load balance, and ensures the stable operation of the distribution network. Specifically, it includes:

[0064] Initialization: Generate an initial solution set or population, and set optimization parameters (including crossover rate, mutation rate, and population size, which are 0.8, 0.1, and 100, respectively);

[0065] Fitness evaluation: Calculate the fitness of each solution. The fitness function is based on the objective function and constraints.

[0066] Select the optimal solution according to the fitness to perform crossover and mutation operations to generate a new solution set or population;

[0067] Repeat the fitness evaluation and selection, crossover, and mutation operations until the convergence condition is met or the maximum number of iterations is reached.

[0068] Optionally, the objective function is expressed as:

[0069]

[0070] Among them, C i represents the cost of adding a charging pile to node i;

[0071] New Chargers i Indicates the number of newly added charging piles at node i;

[0072] λ represents the load balancing weight coefficient;

[0073] N represents the total number of nodes;

[0074] T represents the total number of time periods;

[0075] Adjusted Load i,t represents the adjusted total load of node i in time period t;

[0076] Predicted EV Load i,t represents the electric vehicle charging load predicted by node i in time period t;

[0077] Dynamic Capacity i,t represents the dynamic carrying capacity of node i in time period t;

[0078] The constraints are expressed as:

[0079] Restrictions on the number of charging piles:

[0080] Load balancing constraints:

[0081]

[0082] Planning cost constraints: Among them, Budget is the total budget of the plan.

[0083] Beneficial effects of the present invention:

[0084] The present invention collects electric vehicle charging demand data and urban distribution network load data, and based on these data, uses dynamic load adjustment coefficients and fluctuation coefficients to accurately evaluate the dynamic carrying capacity of each node in the distribution network. By comprehensively considering charging demand and load balance, an optimization algorithm is used to solve the optimal layout plan of new charging piles, which effectively avoids the overload phenomenon of the distribution network and improves the stability and reliability of the power grid operation. At the same time, through a reasonable layout of charging piles, the charging infrastructure resources are maximized, resource waste is avoided, and the utilization efficiency of the charging piles is improved.

[0085] The present invention comprehensively considers planning cost and load balance in the optimization objective function, and introduces a load balance weight coefficient to ensure that the planning scheme minimizes the cost of adding new charging piles while meeting the charging demand and load balance. The optimization algorithm gradually approaches the optimal solution through iterative optimization, and finally obtains an economical and efficient charging pile layout scheme, which not only meets the charging demand of electric vehicles, but also effectively controls the planning cost, achieving a balance between economy and feasibility.

[0086] The present invention introduces a dynamic load adjustment coefficient and a load fluctuation coefficient. Through a comprehensive analysis of factors such as seasonality and special events, the load evaluation method is dynamically adjusted to accurately reflect the actual load conditions of the distribution network under different conditions. By utilizing a dynamically adjusted load calculation method and a collaborative optimization model, an efficient collaborative planning method based on historical data and predicted data is proposed. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0088] Figure 1 A schematic diagram of a planning method flow chart of an embodiment of the present invention;

[0089] Figure 2 Schematic diagram of a collaborative optimization model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0090] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0091] It should be noted that the references to "one embodiment", "an embodiment", "an exemplary embodiment", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0092] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0093] like Figure 1-Figure 2 As shown, the coordinated planning method of electric vehicle charging infrastructure and urban distribution network includes the following steps:

[0094] S1, data collection: collect the historical electric vehicle charging demand data of existing charging piles and the load data of the urban distribution network, including the frequency of use of charging piles, charging time, charging power, and the load conditions of each node of the distribution network. Data collection also includes data preprocessing, which includes normalizing the collected data and completing missing data.

[0095] Normalization: Use the minimum and maximum normalization method to scale the data to the range of [0, 1], expressed as: Among them, x is the original data value, x min and x max are the minimum and maximum values ​​in the data set, respectively, and x′ is the normalized value;

[0096] Missing data is filled by interpolation: for continuous data, linear interpolation or polynomial interpolation is used to fill missing data;

[0097] S2, demand forecasting: Based on the preprocessed data, the machine learning algorithm is used to predict the future charging demand of electric vehicles and generate forecast results covering the charging demand and distribution in different time periods;

[0098] S3, distribution network carrying capacity analysis: Based on the distribution network load data and electric vehicle charging demand forecast results, analyze the carrying capacity of each node in the distribution network and identify overloaded nodes and time periods;

[0099] S4, planning of new charging infrastructure: based on the results of the distribution network carrying capacity analysis, preliminarily determine the layout of new charging piles, including the optimal layout location and quantity, giving priority to areas with low distribution network load and high demand for electric vehicle charging;

[0100] S5, Collaborative Optimization: Establish a collaborative optimization model for charging infrastructure planning and distribution network load balancing, comprehensively consider charging demand, distribution network load and planning cost, use optimization algorithm to solve the optimal solution, and obtain the optimal layout plan for new charging piles.

[0101] The collection of charging demand data for electric vehicles includes:

[0102] Charging pile usage frequency: By collecting statistics on the usage records of each charging pile, the number of times each charging pile is used in a specific time period can be obtained;

[0103] Charging time: record the start and end time of each charge, calculate the charging duration, and summarize the duration data;

[0104] Charging power: monitor and record the charging power changes during each charging process, and summarize the average charging power and peak power of each charging pile in different time periods;

[0105] The collection of urban distribution network load data includes:

[0106] Load conditions at each node of the distribution network: The distribution network monitoring system collects load data of each node of the distribution network in real time, including voltage, current, and power factor, and records load changes in different time periods.

[0107] The machine learning algorithm in S2 uses a decision tree regression model. S2 specifically includes:

[0108] S21, extracting features from the preprocessed data, including time features (hours, weeks, months), spatial features (charging station locations), and historical charging demand features (average charging demand in the past week), and encoding the features to ensure that they are suitable for machine learning algorithms, such as converting time features into periodic features;

[0109] S22, dividing the preprocessed data set into a training set and a test set in a ratio of 7:3 or 8:2 to ensure the effectiveness of model training and the accuracy of prediction results;

[0110] S23, select the decision tree regression model for demand forecasting. The decision tree regression model divides the data set into several subsets, builds a tree structure in each subset, and finds the best split point so that the data on each leaf node has the lowest mean square error (MSE). The goal of decision tree regression is to minimize the mean square error, which is expressed as:

[0111] Among them, y i is the actual value, is the predicted value, n is the number of samples;

[0112] S24, the test set is used to evaluate the trained decision tree regression model, and the evaluation index is used to evaluate the prediction performance of the model. The evaluation is expressed as:

[0113]

[0114] S25, based on the evaluation results, the model is optimized, including adjusting the depth of the decision tree, the minimum number of sample splits, and the minimum number of sample leaf nodes to prevent overfitting or underfitting and improve the prediction accuracy of the model.

[0115] The decision tree regression model specifically includes:

[0116] Select the split point: At each node, select a feature and its corresponding split point to divide the data set into two subsets so that the samples in the split subsets are more similar. The criterion for selecting the split point is to minimize the mean square error after the split. The standard formula for setting the split point is: Among them, n L and n R are the number of samples of the left child node and the right child node respectively, and are the average values ​​of the left and right child nodes, y i and jare the actual values ​​of the left child node and the right child node respectively;

[0117] Recursive splitting: Repeatedly select split points for each subset until the preset stopping condition is reached (such as maximum tree depth, minimum number of sample splits, minimum number of sample leaf nodes, etc.).

[0118] The distribution network carrying capacity analysis in S3 specifically includes:

[0119] S31, integrating the collected distribution network load data with the electric vehicle charging demand forecast results to form a comprehensive database, which includes historical load data of each node in the distribution network and predicted future charging demand data;

[0120] S32, node load dynamic adjustment: Based on historical load data and predicted charging demand data, dynamic load adjustment coefficients are introduced, including basic load dynamic adjustment coefficient α t , Dynamic adjustment coefficient of electric vehicle charging load β t , to reflect the impact of distribution network on load in different seasons and special events (holidays, large-scale activities);

[0121] S33, dynamic analysis of node carrying capacity: analyze the dynamic adjustment load of each node, consider the volatility and extreme conditions of the node load, obtain the load fluctuation range and frequency of each node through historical data analysis, and introduce the fluctuation coefficient γ to reflect the impact of node load fluctuation;

[0122] S34, Dynamic identification of overloaded nodes and time periods: Dynamic analysis of the adjusted node load, combined with the fluctuation coefficient γ i and adjust the load to determine whether the node is at risk of overload in certain periods of time. The dynamic overload determination formula is:

[0123]

[0124] in;

[0125] Adjusted_Load i,t represents the adjusted total load of node i in time period t;

[0126] Dynamic_Capacity i,t represents the dynamic carrying capacity of node i in time period t;

[0127] Overload i,t Indicates the result of determining whether node i is overloaded in time period t.

[0128] The dynamic adjustment of node load is expressed as:

[0129] Adjusted_Load i,t=Base_Load i,t ·(1+α t )+Predicted_EV_Load i,t ·(1+β t ),in;

[0130] Adjusted_Load i,t represents the adjusted total load of node i in time period t;

[0131] Base_Load i,t represents the basic load of node i in time period t;

[0132] Predicted_EV_Load i,t represents the electric vehicle charging load predicted by node i in time period t;

[0133] α t It represents the dynamic adjustment coefficient of the base load in time period t, reflecting the impact of seasonality and special events on the base load;

[0134] β t It represents the dynamic adjustment coefficient of electric vehicle charging load in time period t, reflecting the impact of seasonality and special events on charging load.

[0135] The fluctuation coefficient γ is introduced to reflect the influence of node load fluctuation, which is expressed as:

[0136] Dynamic_Capacity i,t =Max_Capacity i ·(1-γ i ),in;

[0137] Dynamic_Capacity i,t represents the dynamic carrying capacity of node i in time period t;

[0138] Max_Capacity represents the maximum carrying capacity of node i;

[0139] γ i It represents the load fluctuation coefficient of node i, reflecting the safety margin within the node load fluctuation range.

[0140] αt and βt are determined as follows:

[0141] 1. Seasonal adjustment:

[0142]

[0143] in;

[0144] The average base load for a particular season.

[0145] Average base load throughout the year.

[0146] Average electric vehicle charging load for a specific case.

[0147] Average electric vehicle charging load throughout the year.

[0148] 2. Special event adjustments:

[0149]

[0150] in;

[0151] AverageLoadevent: The average base load during a special event.

[0152] Normal_Load: Basic load under normal circumstances.

[0153] Average_EV_Load_event: Average electric vehicle charging load during the special event.

[0154] Normal_EV_Load: Electric vehicle charging load under normal conditions.

[0155] 3. Comprehensive adjustment coefficient:

[0156] α t =α season +α event ;

[0157] β t =β season +β event ;

[0158] γ i Determination of:

[0159] σ i : The standard deviation of load fluctuation at node i.

[0160] ΔMax Load i : Maximum load fluctuation value of node i.

[0161] Max Capacity i : The maximum carrying capacity of node i.

[0162] γ i : Load fluctuation coefficient of node i.

[0163] S4 specifically includes:

[0164] Node priority evaluation formula:

[0165] in;

[0166] Priority i represents the priority of node i;

[0167] represents the electric vehicle charging load predicted by node i in time period t;

[0168] Dynamic Capacity i,t represents the dynamic carrying capacity of node i in time period t;

[0169] represents the adjusted total load of node i in time period t;

[0170] According to the node priority evaluation results, high-priority nodes are selected as the layout locations of new charging piles, and priority is given to nodes with high dynamic carrying capacity and low load to ensure the reliability and effectiveness of the new charging piles.

[0171] Determine the number of charging piles: Based on the charging demand forecast results and the dynamic carrying capacity of the node, determine the number of new charging piles for each priority node:

[0172] Among them, New Chargers i Indicates the number of charging piles that need to be added to node i;

[0173] Average Charger Capacity indicates the average charging capacity of a single charging pile.

[0174] The collaborative optimization in S5 uses the preliminarily determined layout of new charging piles, including the optimal layout location and quantity, as the input of the collaborative optimization model to ensure that the optimization model is carried out on the basis of the preliminary plan, taking into account the charging demand, distribution network load and planning cost. The collaborative optimization model includes defining the objective function, constraints and optimization algorithm. The optimization algorithm is used to solve the collaborative optimization model to obtain the optimal layout plan of new charging piles. The optimal plan minimizes the planning cost while meeting the charging demand and load balance, and ensures the stable operation of the distribution network. Specifically, it includes:

[0175] Initialization: Generate an initial solution set or population, and set optimization parameters (including crossover rate, mutation rate, and population size, which are 0.8, 0.1, and 100, respectively);

[0176] Fitness evaluation: Calculate the fitness of each solution. The fitness function is based on the objective function and constraints.

[0177] Select the optimal solution according to the fitness to perform crossover and mutation operations to generate a new solution set or population;

[0178] Repeat the fitness evaluation and selection, crossover, and mutation operations until the convergence condition is met or the maximum number of iterations is reached.

[0179] The objective function is expressed as:

[0180]

[0181] Among them, C i represents the cost of adding a new charging pile at node i;

[0182] New Chargers i Indicates the number of newly added charging piles at node i;

[0183] λ represents the load balancing weight coefficient;

[0184] N represents the total number of nodes;

[0185] T represents the total number of time periods;

[0186] Adjusted Load i,t represents the adjusted total load of node i in time period t;

[0187] Predicted EV Load i,t represents the electric vehicle charging load predicted by node i in time period t;

[0188] Dynamic Capacity i,t represents the dynamic carrying capacity of node i in time period t;

[0189] The constraints are expressed as:

[0190] Restrictions on the number of charging piles:

[0191] Load balancing constraints:

[0192]

[0193] Planning cost constraints: Among them, Budget is the total budget of the plan.

[0194] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0195] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for collaborative planning of electric vehicle charging infrastructure and urban distribution network, characterized in that: The following steps are involved: S1, data collection: collect historical electric vehicle charging demand data of existing charging piles and urban distribution network load data, including charging pile usage frequency, charging time, charging power, and load conditions of each node in the distribution network. The data collection also includes data preprocessing, which includes normalizing the collected data and completing missing data. S2, demand forecasting: Based on the preprocessed data, the machine learning algorithm is used to predict the future charging demand of electric vehicles and generate forecast results covering the charging demand and distribution in different time periods; S3, distribution network carrying capacity analysis: Based on the distribution network load data and electric vehicle charging demand forecast results, analyze the carrying capacity of each node in the distribution network and identify overloaded nodes and time periods; S4, planning of new charging infrastructure: based on the results of the distribution network carrying capacity analysis, preliminarily determine the layout of new charging piles, including the optimal layout location and quantity; S5, Collaborative Optimization: Establish a collaborative optimization model for charging infrastructure planning and distribution network load balancing, comprehensively consider charging demand, distribution network load and planning cost, use optimization algorithm to solve the optimal solution, and obtain the optimal layout plan for new charging piles.

2. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 1, characterized in that: The collection of historical electric vehicle charging demand data includes: Charging pile usage frequency: By collecting statistics on the usage records of each charging pile, the number of times each charging pile is used in a specific time period can be obtained; Charging time: record the start and end time of each charge, calculate the charging duration, and summarize the duration data; Charging power: monitor and record the charging power changes during each charging process, and summarize the average charging power and peak power of each charging pile in different time periods; The collection of urban distribution network load data includes: Load conditions at each node of the distribution network: The distribution network monitoring system collects load data of each node of the distribution network in real time, including voltage, current, and power factor, and records load changes in different time periods.

3. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 1, characterized in that: The machine learning algorithm in S2 adopts a decision tree regression model, and S2 specifically includes: S21, extracting features from the preprocessed data, including time features, spatial features, and historical charging demand features, and encoding the features; S22, dividing the preprocessed data set into a training set and a test set; S23, select the decision tree regression model for demand forecasting. The decision tree regression model divides the data set into several subsets, builds a tree structure in each subset, and finds the best split point so that the data on each leaf node has the lowest mean square error (, the goal of decision tree regression is to minimize the mean square error, which is expressed as: Among them, y i is the actual value, is the predicted value, n is the number of samples; S24, use the test set to evaluate the trained decision tree regression model, and use the evaluation index to evaluate the prediction performance of the model; S25, based on the evaluation results, the model is optimized, including adjusting the depth of the decision tree, the minimum number of sample splits, and the minimum number of sample leaf nodes to prevent overfitting or underfitting.

4. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 3 is characterized in that: The decision tree regression model specifically includes: Select the split point: At each node, select a feature and its corresponding split point to divide the data set into two subsets so that the samples in the split subsets are more similar. The criterion for selecting the split point is to minimize the mean square error after the split. The standard formula for setting the split point is: Among them, n L and n R are the number of samples of the left child node and the right child node respectively, and are the average values ​​of the left and right child nodes, y i and j are the actual values ​​of the left child node and the right child node respectively; Recursive splitting: Repeatedly select split points for each subset until a preset stopping condition is reached.

5. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 1, characterized in that: The distribution network carrying capacity analysis in S3 specifically includes: S31, integrating the collected distribution network load data with the electric vehicle charging demand forecast results to form a comprehensive database, which includes historical load data of each node in the distribution network and predicted future charging demand data; S32, node load dynamic adjustment: Based on historical load data and predicted charging demand data, dynamic load adjustment coefficients are introduced, including basic load dynamic adjustment coefficient α t , Dynamic adjustment coefficient of electric vehicle charging load β t , to reflect the impact of distribution network on load in different seasons and special events; S33, dynamic analysis of node carrying capacity: analyze the dynamic adjustment load of each node, consider the volatility and extreme conditions of the node load, obtain the load fluctuation range and frequency of each node through historical data analysis, and introduce the fluctuation coefficient γ to reflect the impact of node load fluctuation; S34, Dynamic identification of overloaded nodes and time periods: Dynamic analysis of the adjusted node load, combined with the fluctuation coefficient γ i and adjust the load to determine whether the node is at risk of overload in certain periods of time. The dynamic overload determination formula is: in; Adjusted_Load i,t represents the adjusted total load of node i in time period t; Dynamic_Capacity i,t represents the dynamic carrying capacity of node i in time period t; Overload i,t Indicates the result of determining whether node i is overloaded in time period t.

6. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 5, characterized in that: The node load dynamic adjustment is expressed as: Adjusted_Load i,t =Base_Load i,t ·(1+α t )+Predicted_EV_Load i,t · (1+β t ),in; Adjusted_Load i,t represents the adjusted total load of node i in time period t; Base_Load i,t represents the basic load of node i in time period t; Predicted_EV_Load i,t represents the electric vehicle charging load predicted by node i in time period t; α t It represents the dynamic adjustment coefficient of the base load in time period t, reflecting the impact of seasonality and special events on the base load; β t It represents the dynamic adjustment coefficient of electric vehicle charging load in time period t, reflecting the impact of seasonality and special events on charging load.

7. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 5, characterized in that: The introduction of the fluctuation coefficient γ to reflect the influence of node load fluctuation is expressed as: Dynamic_Capacity i,t = Max_Capacity i ·(1 - γ i ), where; Dynamic_Capacity i,t represents the dynamic carrying capacity of node i in time period t; Max_Capacity represents the maximum carrying capacity of node i; γ i It represents the load fluctuation coefficient of node i, reflecting the safety margin within the node load fluctuation range.

8. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 1, characterized in that: The S4 specifically includes: Node priority evaluation formula: in; Priority i Indicates the priority of node i; represents the electric vehicle charging load predicted by node i in time period t; Dynamic Capacity i,t represents the dynamic carrying capacity of node i in time period t; represents the adjusted total load of node i in time period t; According to the node priority evaluation results, select high-priority nodes as the locations for the new charging piles; Determine the number of charging piles: Based on the charging demand forecast results and the dynamic carrying capacity of the node, determine the number of new charging piles for each priority node: in, New Chargers i Indicates the number of charging piles that need to be added to node i; Average Charger Capacity indicates the average charging capacity of a single charging pile.

9. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 1, characterized in that: The collaborative optimization in S5 uses the newly determined layout of charging piles, including the optimal layout position and quantity, as the input of the collaborative optimization model, comprehensively considering the charging demand, distribution network load and planning cost. The collaborative optimization model includes defining the objective function, constraint conditions and optimization algorithm. The optimization algorithm is used to solve the collaborative optimization model to obtain the optimal layout plan of newly added charging piles. The optimal plan minimizes the planning cost while meeting the charging demand and load balance, and ensures the stable operation of the distribution network. Specifically, it includes: Initialization: Generate an initial solution set or population, and set optimization parameters (including crossover rate, mutation rate, and population size, which are 0.8, 0.1, and 100, respectively); Fitness evaluation: Calculate the fitness of each solution. The fitness function is based on the objective function and constraints. Select the optimal solution according to the fitness to perform crossover and mutation operations to generate a new solution set or population; Repeat the fitness evaluation and selection, crossover, and mutation operations until the convergence condition is met or the maximum number of iterations is reached.

10. The method for collaborative planning of electric vehicle charging infrastructure and urban distribution network according to claim 9, characterized in that: The objective function is expressed as: Among them, C i represents the cost of adding a new charging pile at node i; New Chargers i Indicates the number of newly added charging piles at node i; λ represents the load balancing weight coefficient; N represents the total number of nodes; T represents the total number of time periods; Adjusted Load i,t represents the adjusted total load of node i in time period t; Predicted EV Load i,t represents the electric vehicle charging load predicted by node i in time period t; Dynamic Capacity i,t represents the dynamic carrying capacity of node i in time period t; The constraints are expressed as: Restrictions on the number of charging piles: Load balancing constraints: Planning cost constraints: Among them, Budget is the total budget of the plan.

Images