A Highway Charging Facility Layout Planning Method Considering Range Anxiety

Abstract

This invention provides a highway charging facility layout planning method considering range anxiety, comprising the following steps: Step S1: Generating OD matrix pairs based on user driving data recorded by the highway toll system; Step S2: Updating traffic flow information for each road using road and node models in a segment transmission model based on the OD matrix pairs, and reflecting user charging demand based on traffic flow information; Step S3: Establishing candidate sites, thereby obtaining a pre-selection of candidate site schemes; Step S4: Setting the objective function to minimize charging station operation and maintenance costs and user queuing time, and calculating queuing time; Step S5: Solving using the NSGAⅡ multi-objective algorithm to obtain the Pareto front; Step S6: Selecting the final implementation scheme from the Pareto solution using evidence reasoning, and substituting it into the user range anxiety level formula to verify and improve the scheme. Applying this technical solution can effectively reduce the cost of highway charging station layout planning and user anxiety levels.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle charging facility layout planning technology, and in particular to a highway charging facility layout planning method that takes range anxiety into account. Background Technology

[0002] With the continuous promotion of the new energy vehicle industry by the country, the number of electric vehicle users and charging facilities is increasing, as are their operating hours. This leads to an exponential increase in user data collected at highway toll stations and service areas, exhibiting characteristics of being massive, real-time, dynamic, and diverse. Big data analytics can be used to accurately characterize user charging behavior, accurately assess spatiotemporal charging demand, and achieve dynamic analysis and modeling, providing a data foundation for the rational planning of charging facilities. Therefore, the planning of the charging layout of electric vehicle charging facilities on highways is actually a complex, nonlinear, and strongly coupled multi-objective optimization problem. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide a highway charging facility layout planning method that takes into account range anxiety, so as to effectively reduce the cost of highway charging station layout planning and reduce user anxiety.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a highway charging facility layout planning method considering range anxiety, comprising the following steps:

[0005] Step S1: Generate OD matrix pairs based on user driving data recorded by the highway toll system;

[0006] Step S2: A method based on the OD matrix to update the traffic flow information of each road using the road model and node model in the road segment transmission model, and to reflect the user's charging demand based on the traffic flow information;

[0007] Step S3: Establish candidate sites and impose distance restrictions between charging stations based on users' range anxiety and charging needs, thereby obtaining a pre-selection of candidate site schemes;

[0008] Step S4: Set the objective function to minimize the charging station operation and maintenance cost and the user queuing time, and calculate the queuing time based on the M / M / K queuing theory.

[0009] Step S5: Use the NSGAⅡ multi-objective algorithm to obtain the Pareto front;

[0010] Step S6: Use evidence-based reasoning to select the final implementation plan from the Pareto solution, and substitute it into the user's mileage anxiety formula to verify and improve the plan.

[0011] In a preferred embodiment, dynamic simulation is used to obtain the spatiotemporal distribution of traffic flow based on the OD matrix, including the following steps:

[0012] Step 21: Analyze the National Day data based on the gantry information of the Fujian section of G15 Shenhai Expressway to obtain the average OD matrix pair;

[0013] Step 22: For each section of the highway route, specify the length L. i Blocking density Free flow velocity v i , the maximum traffic capacity of the road Configure parameters;

[0014] Step 23: Calculate the transmission capacity S for each road segment. i (t) and receiving capability R i (t), which is related to the cumulative traffic volume of the upstream and downstream road sections and the parameter settings in step 22;

[0015]

[0016]

[0017] In the formula, L i v i w i and These represent the maximum capacity, upstream inlet location, downstream outlet location, length, free-flow velocity, shock wave reverse propagation velocity, and blockage density of the i-th road, respectively; Δt represents the time variation; min(·) indicates taking the minimum value within the parentheses; Indicates the location at the upstream entrance of the i-th road. Cumulative traffic flow at any given time; This represents the cumulative traffic flow at time t at the upstream entrance of the i-th road;

[0018] Step 24: Establish the node model in the dynamic road segment transmission model, substitute the cumulative traffic flow of each road segment obtained in Step 23 into the node model, and obtain the transfer flow between adjacent road segments according to different node types;

[0019] Step 25: Based on the inflow, outflow, and transfer flow of traffic for each road segment obtained in Steps 23 and 24, update the cumulative inflow and outflow traffic information for each road segment.

[0020]

[0021]

[0022] In the formula, G represents the cumulative traffic flow at time t downstream of line i; ij (t) represents the transfer traffic flow from line i to line j; The cumulative traffic flow at time t upstream of line j;

[0023] Suppose that for any path p, it passes through road segment i and road segment j, and j is a downstream road segment of i, then,

[0024]

[0025]

[0026] In the formula, N p (x i L ,t+Δt) represents the cumulative traffic flow at time t+Δt, which is downstream of line i in path p; The cumulative traffic flow at the upstream position of line i is The corresponding time; δ represents the cumulative traffic flow at time t+Δt, upstream of path p on line j; jp G is a road segment association matrix, where G is 1 if path p contains line j, and 0 otherwise; ij (t) represents the transfer flow that is transferred from line i to line j at time t;

[0027] Step 26: Based on the cumulative traffic flow information, the traffic flow density can be calculated for any point on any road segment. That is, for a road segment i, assuming point x is on the road segment, the transmitting capacity S(x,t) and receiving capacity R(x,t) at that point are respectively

[0028]

[0029]

[0030] In the formula, x, v, w and These represent the upstream position, downstream position, any position on line i, free flow velocity, turbulent reverse propagation velocity, and blockage density, respectively. This refers to the traffic flow information at time t at position x on the i-th route; For the upstream position of the i-th line The cumulative traffic flow at any given moment.

[0031] In a preferred embodiment, step 3 specifically includes:

[0032] Step 31: Calculate the current range anxiety level of electric vehicle users based on the number of charging stations and charging piles;

[0033]

[0034] In the formula, Let N(a,b) represent the range anxiety of EV user i given the current number of charging stations. Assume that N(a,b) follows a normal distribution, where a represents the average range anxiety level of electric vehicle users, i.e., the minimum SOC that electric vehicle users can tolerate when arriving at a charging station, and b is its corresponding variance.

[0035] Step 32: Based on the range anxiety levels of users according to a normal distribution, for confidence interval probabilities of 95.45% and 99.74%, the maximum value in the set of all users' minimum SOC values ​​at arrival is used to limit the distance between charging stations. That is, 95.45% and 99.74% of electric vehicle users will have an SOC value greater than their minimum SOC value at arrival under this distance limit.

[0036] Step 33: Substitute the final candidate results into the range anxiety formula to obtain the change in user range anxiety under this charging station layout scheme, i.e.

[0037]

[0038]

[0039] In the formula, n cs This represents the total number of charging piles; α and β are the corresponding hyperparameters; as the number of charging piles in a charging station increases... The level of range anxiety among users given the current charging station layout; To assess users' mileage anxiety level after implementing the optimal layout plan; B represents the battery capacity of electric vehicle i. r,i This represents the maximum battery capacity required for electric vehicle i to reach the charging station in the layout planning results.

[0040] Step 34: Based on the user's range anxiety level obtained in Step 33, use the maximum value in the lowest SOC set with confidence interval probabilities of 95.45% and 99.74% to calculate the distance limit between charging stations until the user's range anxiety level and the distance between charging stations remain basically unchanged.

[0041] l limit =B r,i *l i (12)

[0042] In the formula, l limit Distance limitations between charging stations; B r,i Let l be the maximum battery capacity required for electric vehicle i to reach the charging station in the layout planning results;i For electric vehicle users of type i, the driving range is 1000 km / h.

[0043] In a preferred embodiment, the construction and operation costs of charging stations and user psychological characteristics are used as objective functions. Different weighted objective functions are adopted according to the actual problem, and the Pareto front is obtained by calculating multiple objective functions through the NSGA II algorithm.

[0044] Step 41: Substitute the number of charging stations to be built, the minimum number of charging piles at each charging station if a charging station is built on each line, the distance limit between charging stations, the traffic flow information at any location and time, and the candidate station sites as parameters.

[0045] Step 42: Preset the NSGAⅡ parameters, with the independent variables being the location of the charging station and the corresponding number of charging piles;

[0046] Step 43: Set the objective function to the construction and operation cost C of the charging station. cost User queuing time T wait , respectively

[0047]

[0048] In the formula, γ represents the annual operation and maintenance parameters; r0 is the discount rate; z is the operating life of the charging station; N cs c is the preset number of charging stations to be built along the route; f The fixed costs of building charging stations; c bc The fixed cost of the charging station; The number of charging piles built for the i-th charging station;

[0049]

[0050] In the formula, This represents the average waiting time of an electric vehicle at the i-th charging station at time t; This indicates the penetration rate of electric vehicles, which is the average proportion of electric vehicles traveling on highways. The probability that an electric vehicle enters the i-th charging station at time t; The electric vehicle traffic flow is represented by τ at time t; τ is the simulation step size.

[0051] Step 44: Obtain the corresponding Pareto front by changing the weights of the objective function according to actual needs.

[0052] In a preferred embodiment, the final implementation scheme is selected from the Pareto solution of the highway electric vehicle charging station layout planning based on evidence-based reasoning according to the actual problem;

[0053] Step 51: Based on Step 44, obtain the Pareto candidate solution set A for the layout planning of electric vehicle charging stations on highways. Each candidate solution includes the location information of the charging stations to be built and the corresponding number of charging piles; and determine the evaluation index set E and the evaluation level set H, respectively.

[0054] A = {a1, a2, ..., a} j ,…a M}(j=1,2,…M) (15)

[0055] E = {e1,e2,…,e} i ,…,e L}(i=1,2,…,L) (16)

[0056] H = {H1,H2,…,H} n ,…H N}(n=1,2,…,N) (17)

[0057] In the formula, a1, a2, ..., a j ,…a M Candidate solutions for site selection and sizing of electric vehicle charging stations in the Pareto frontier; e1, e2, ..., e i ,…,e L To evaluate different metrics, i.e., the corresponding multi-objective optimization functions: charging station construction and operation costs, total user queuing time; H1, H2, ..., H n ,…H N The rating is usually described using natural language, such as {poor, poor, average, good, excellent}.

[0058] Step 52: For each candidate solution for charging station layout planning, evaluate the confidence level of each element in the evaluation index set at different evaluation levels to obtain the confidence vector, i.e.

[0059] S(e i (a j ))={(H n ,β n,i (a j )), n=1,2,…,N; i=1,2,…,L; j=1,2,…,M} (18)

[0060] In the formula, H n For the nth evaluation level; a j β is the j-th candidate solution in the candidate solution for the site selection and layout planning of electric vehicle charging stations. n,i (a j ) is a jThe confidence level for the i-th evaluation index system and the n-th evaluation level represents the decision-maker's level of understanding of the evaluation index, and satisfies the following conditions:

[0061] 0≤β n,i (a j )≤1 (19)

[0062]

[0063] Step 53: Fuse multiple confidence evaluation vectors to obtain the overall evaluation of Pareto candidate solutions for electric vehicle charging station layout planning; first, based on the assessed confidence β... n,i Obtain the basic credibility m n,i It is then divided into two parts; secondly, the uncertainty basic confidence level m is calculated. H,i It is also divided into two parts; then it will be evaluated in level H. n The confidence scores of the first i attributes are combined to obtain the total basic confidence score m. n,I(i) Ultimately, the evaluation level was H. n General's multi-attribute set {e1,e2,…,e i ,…e L The confidence scores of the candidate schemes for electric vehicle charging station layout planning are fused to obtain the confidence vector S(y) in the evaluation level set.

[0064] m n,i =ω i β n,i (twenty one)

[0065]

[0066]

[0067]

[0068]

[0069] {H n}:m n,I(i+1) =K I(i+1) [m n,J(i) m n,i+1 +m H,I(i) m n,i+1 +m n,I(i) m H,i+1 (26)

[0070]

[0071] Where m n,I(i) The total basic confidence number not assigned to the first i indicators is decomposed into:

[0072]

[0073] in

[0074]

[0075] In the formula, y represents the candidate scheme, and β n (y) represents the total confidence level of scheme y at the nth evaluation level, while β H (y) represents the overall uncertainty confidence level of the decision-maker in being unable to assess option y, β n (y) and β H The expression for (y) is

[0076]

[0077] Step 54: Map the confidence distribution of electric vehicle charging stations to utility values ​​through utility analysis, compare the average utility values ​​of different candidate solutions, and select the final candidate solution;

[0078] First, the confidence distribution vector is mapped to utility values ​​through utility analysis:

[0079]

[0080] Where u(H) n ) is the evaluation level H n The utility value is such that the higher the rating level, the greater the corresponding utility value, i.e., u(H). n+1 )>u(H n (H1 < H2 < ... H) n <H n+1 <...H N Define the formulas for maximum, minimum, and average utility values, i.e.

[0081]

[0082] Ultimately, the average utility value was chosen as the criterion for evaluating the site selection and sizing of candidate solutions for electric vehicle charging stations.

[0083]

[0084] Compared with existing technologies, this invention has the following advantages: This invention provides a highway charging facility layout planning method that considers range anxiety. First, a road segment transmission model can be used to dynamically simulate traffic flow, and the NSGA II algorithm is applied to solve the objective function using a queuing theory model. Then, reflecting the distance limitations between charging stations by the user's range anxiety level can effectively reduce the user's psychological anxiety; and evidence-based reasoning is used to select the final implementation scheme from the Pareto solution. Attached Figure Description

[0085] Figure 1 This is a flowchart of a method for planning the layout of highway charging facilities considering range anxiety, which is a preferred embodiment of the present invention. Detailed Implementation

[0086] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0087] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. 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 application pertains.

[0088] 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 according to this application; as used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise; furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0089] S11: Generate OD matrix pairs based on user driving data recorded by the highway toll system;

[0090] S12: Based on the OD matrix, update the traffic flow information of each road using the road model and node model in the road segment transmission model, and reflect the user's charging demand based on the traffic flow information;

[0091] S13: Establish candidate sites and limit the distance between charging stations based on users' range anxiety and charging needs, thereby obtaining a pre-selection of candidate site schemes;

[0092] S14: Set the objective function to minimize the charging station operation and maintenance cost and the user queuing time, and calculate the queuing time based on the M / M / K queuing theory.

[0093] S15: The Pareto front is obtained by solving the problem using the NSGAⅡ multi-objective algorithm;

[0094] S16: Use evidence-based reasoning to select the final implementation scheme from the Pareto solution, and substitute it into the user's mileage anxiety formula to verify and improve the scheme.

[0095] Specifically:

[0096] 1. Generate OD matrix pairs based on highway user driving data and after data processing.

[0097] Second: Update the traffic flow information for each road based on the OD matrix and the road and node models in the road segment transmission model. This mainly includes the following steps:

[0098] Step 21: Analyze the National Day data based on the gantry information of the Fujian section of G15 Shenhai Expressway to obtain the average OD matrix pair;

[0099] Step 22: For each section of the highway route, specify the length L. i Blocking density Free flow velocity v i , the maximum traffic capacity of the road Configure parameters;

[0100] Step 23: Calculate the transmission and reception capacity of each road segment, which is related to the cumulative traffic volume of the upstream and downstream road segments and the parameter settings in Step 22;

[0101]

[0102]

[0103] L i v i , These are the maximum capacity, upstream inlet location, downstream outlet location, length, free-flow velocity, and congestion density of the i-th road, respectively. Indicates the location at the upstream entrance of the i-th road. The cumulative traffic flow at any given moment.

[0104] Step 24: Establish the node model in the dynamic road segment transmission model, substitute the cumulative traffic flow of each road segment obtained in Step 23 into the node model, and obtain the transfer flow between adjacent road segments according to different node types;

[0105] For the source node, its traffic is

[0106]

[0107] In the formula, R represents the cumulative traffic flow at time t upstream of road segment j; j (t) represents the received stream of segment j.

[0108] For the connected node, the traffic is

[0109] G ij (t)=min{S i (t),R j (t)} (4)

[0110] In the formula, S i(t) represents the transmission flow of segment i; R j (t) represents the received stream of segment j.

[0111] For the splitter node, its traffic is

[0112]

[0113] In the formula, S ij (t) represents the transmission stream sent from segment i to j; R ij′ (t) represents the received stream sent from segment i to j′; S ij (t) represents the transmission stream from segment i to j′.

[0114] For the merging node, its flow rate is

[0115] G ij (t)=min{S ij (t),p ij R j (t)} (6)

[0116] In the formula S ij (t) represents the transmission stream sent from segment i to j; R j (t) represents the received stream of segment j. ij The percentage of vehicles from road segment i to road segment j, satisfying

[0117] For the terminal node, its flow is

[0118] G i (t)=S i (t) (7)

[0119] Step 25: Based on the inflow, outflow, and transfer flow of traffic for each road segment obtained in Steps 23 and 24, update the cumulative inflow and outflow traffic information for each road segment;

[0120]

[0121]

[0122] In the formula, G represents the cumulative traffic flow at time t downstream of line i; ij (t) represents the transfer traffic flow from line i to line j; This represents the cumulative traffic flow at time t upstream of line j.

[0123] Suppose that for any path p, it passes through road segment i and road segment j, and j is a downstream road segment of i, then,

[0124]

[0125]

[0126] In the formula, Let be the cumulative traffic flow at time t+Δt, which is downstream of line i in path p; The cumulative traffic flow at the upstream position of line i is The corresponding time; δ represents the cumulative traffic flow at time t+Δt, upstream of path p on line j; jp G is a road segment association matrix, where G is 1 if path p contains line j, and 0 otherwise; ij (t) represents the transfer flow that is transferred from line i to line j at time t.

[0127] Step 26: Based on the cumulative traffic flow information, the traffic flow density can be calculated for any point on any road segment. That is, for a road segment i, assuming point x is on the road segment, the transmitting capacity S(x,t) and receiving capacity R(x,t) at that point are respectively

[0128]

[0129]

[0130] In the formula, x, v, w These represent the upstream position, downstream position, any position on line i, free flow velocity, turbulent reverse propagation velocity, and blockage density, respectively. This refers to the traffic flow information at time t at position x on the i-th route; For the upstream position of the i-th line The cumulative traffic flow at any given moment.

[0131] Third: Establish candidate sites and limit the distance between charging stations based on users' range anxiety and charging needs, thereby obtaining a pre-selection of candidate site schemes.

[0132] Step 31: Calculate the current range anxiety level of electric vehicle users based on the number of charging stations and charging piles;

[0133]

[0134] Let N(a,b) represent the range anxiety of EV user i given the current number of charging stations. Assume that N(a,b) follows a normal distribution, where a represents the average range anxiety level of electric vehicle users, i.e., the minimum SOC that electric vehicle users can tolerate when arriving at a charging station, and b is its corresponding variance.

[0135] Step 32: Based on the normally distributed mileage anxiety levels of users, assign a confidence interval probability of 95.45% (corresponding to the horizontal axis interval is...) The confidence interval probability is 99.74% (corresponding to the horizontal axis interval is...). The distance between charging stations is limited by the maximum value in the set of minimum SOC values ​​for all users at the station. That is, 95.45% and 99.74% of electric vehicle users have a SOC value greater than their minimum SOC value at the station when they arrive at the station under this distance limit.

[0136] Step 33: Substitute the final candidate results selected in Step 54 into the range anxiety formula to obtain the change in user range anxiety under this charging station layout scheme, i.e.

[0137]

[0138]

[0139] In the formula, n cs This represents the total number of charging piles; α and β are the corresponding fixed parameters; as the number of charging piles in the charging station increases... The level of range anxiety among users given the current charging station layout; To assess users' mileage anxiety level after implementing the optimal layout plan; B represents the battery capacity of electric vehicle i. r,i This represents the maximum battery capacity required for electric vehicle i to reach the charging station in the layout planning results.

[0140] Step 34: Based on the user's range anxiety level obtained in Step 33, use the maximum value in the lowest SOC set with confidence interval probabilities of 95.45% and 99.74% to calculate the distance limit between charging stations, and then substitute it into Step 41 until the user's range anxiety level and the distance between charging stations basically do not change.

[0141] l limit =B r,i *l i (17)

[0142] In the formula, l limit Distance limitations between charging stations; B r,i This represents the maximum battery capacity required for electric vehicle i to reach the charging station in the layout planning results. i For the driving range of electric vehicle users of type i;

[0143] Fourth: The construction and operation costs of charging stations and user psychological characteristics are used as objective functions. Different weights are applied to objective functions based on the actual problem, and the NSGA II algorithm is used to calculate the multi-objective function, resulting in the Pareto front. This mainly includes the following steps:

[0144] Step 41: Substitute the number of charging stations to be built, the minimum number of charging piles at each charging station if a charging station is built on each line, the distance limit between charging stations, the traffic flow information at any location and time, and the candidate station sites as parameters.

[0145] Step 42: NSGA II introduces fast non-dominated sorting, congestion degree and congestion distance, and an elite strategy, which can reduce the complexity of non-dominated sorting genetic algorithms. The NSGA II parameters are pre-set, with independent variables being the location of charging stations and the corresponding number of charging piles. The objective function is the construction and maintenance cost of charging stations and the total charging queuing time for electric vehicle users. The constraints are the distance between charging stations and the minimum number of charging piles required for each line to ensure a service intensity greater than 1.0. The parameters are ParetoFraction = 0.3, PopulationSize = 200, MaxGenerations = 500, StallGenLimit = 100, and TolFun = 1e-10.

[0146] Step 43: Set the objective function to the construction and operation cost of charging stations and the user queuing time, respectively.

[0147]

[0148] Where γ is the annual operation and maintenance parameter, which is taken as 3% in this paper; r0 is the discount rate; z is the operating years of the charging station; N cs c is the preset number of charging stations to be built along the route; f The fixed costs of building charging stations; c bc The fixed cost of the charging station; The number of charging piles built for the i-th charging station.

[0149]

[0150] In the formula, This represents the average waiting time of an electric vehicle at the i-th charging station at time t; This indicates the penetration rate of electric vehicles, which is the average proportion of electric vehicles traveling on highways. The probability that an electric vehicle enters the i-th charging station at time t; q(x i ,t) represents the traffic flow of electric vehicles at time t; τ is the simulation step size.

[0151] Step 44: Obtain the corresponding Pareto front by changing the weights of the objective function according to actual needs.

[0152] Fifth: Using evidence-based reasoning, select the final implementation scheme from the Pareto solution of the highway electric vehicle charging station layout plan, and re-substitute the result of the charging station layout plan into the user's range anxiety for calculation, until the final result caused by the user's range anxiety basically no longer changes.

[0153] Step 51: Based on the Pareto candidate solution set A obtained in Step 44, each candidate solution includes the location information of the charging station to be built and the corresponding number of charging piles; and determine the evaluation index set E and the evaluation level set H, respectively.

[0154] A = {a1, a2, ..., a} j ,…a M}(j=1,2,…M) (20)

[0155] E = {e1,e2,…,e} i ,…,e L}(i=1,2,…,L) (21)

[0156] H = {H1,H2,…,H} n ,…H N}(n=1,2,…,N) (22)

[0157] In the formula, a1, a2, ..., a j ,…a M Candidate solutions for site selection and sizing of electric vehicle charging stations in the Pareto frontier; e1, e2, ..., e i ,…,e L To evaluate different metrics, i.e., the corresponding multi-objective optimization functions: charging station construction and operation costs, total user queuing time; H1, H2, ..., H n ,…H N The rating is usually described using natural language, such as {poor, poor, average, good, excellent}.

[0158] Step 52: For each candidate solution for charging station layout planning, evaluate the confidence level of each element in the evaluation index set at different evaluation levels to obtain the confidence vector, i.e.

[0159] S(e i (a j ))={(H n ,β n,i (a j )), n=1,2,…,N; i=1,2,…,L; j=1,2,…,M}(23)

[0160] In the formula, H n For the nth evaluation level; a j β is the j-th candidate solution in the candidate solution for the site selection and layout planning of electric vehicle charging stations. n,i (a j ) is a j The confidence level for the i-th evaluation index system and the n-th evaluation level represents the decision-maker's level of understanding of the evaluation index, and satisfies the following conditions:

[0161] 0≤β n,i (a j )≤1 (24)

[0162]

[0163] Step 53: Fuse multiple confidence evaluation vectors to obtain the overall evaluation of Pareto candidate solutions for electric vehicle charging station layout planning. First, based on the assessed confidence level β... n,i Obtain the basic credibility m n,i It is then divided into two parts; secondly, the uncertainty basic confidence level m is calculated. H,i It is also divided into two parts; then it will be evaluated in level H. n The confidence scores of the first i attributes are combined to obtain the total basic confidence score m. n,I(i) Ultimately, the evaluation level was H. n General's multi-attribute set {e1,e2,…,e i ,…e L The confidence scores of the candidate solutions are fused to obtain the confidence vector S(y) of the candidate solutions in the evaluation level set.

[0164] m n,i =ω i β n,i (26)

[0165]

[0166]

[0167]

[0168]

[0169] {H n}:m n,I(i+1) =K I(i+1) [m n,J(i) m n,i+1 +m H,I(i) m n,i+1 +m n,I(i) m H,i+1(31)

[0170]

[0171] Where m n,I(i) The total basic confidence number, which is not assigned to the first i indicators, can be decomposed into:

[0172]

[0173] in

[0174]

[0175] In the formula, y represents the candidate scheme, and β n (y) represents the total confidence level of scheme y at the nth evaluation level, while β H (y) represents the overall uncertainty confidence level of the decision-maker in being unable to assess option y, β n (y) and β H The expression for (y) is

[0176]

[0177] Step 54: Map the confidence distribution to utility values ​​through utility analysis, compare the average utility values ​​of different candidate solutions, and select the final candidate solution.

[0178] First, the confidence distribution vector is mapped to utility values ​​through utility analysis:

[0179]

[0180] Where u(H) n ) is the evaluation level H n The utility value is such that the higher the rating level, the greater the corresponding utility value, i.e., u(H). n+1 )>u(H n (H1 < H2 < ... H) n <H n+1 <...H N Define the formulas for maximum, minimum, and average utility values, i.e.

[0181]

[0182] Ultimately, the average utility value was chosen as the criterion for evaluating candidate solutions.

[0183]

[0184] This invention provides a highway charging facility layout planning method that considers range anxiety. It generates OD matrix pairs based on user driving data recorded by the highway toll system, and uses a road segment transmission model to obtain traffic flow information for each road based on these OD matrix pairs, thereby reflecting user charging demand. Candidate site selection is established, and a preliminary selection of candidate site schemes is performed based on user range anxiety levels, user charging demand, and user charging preferences. The objective function is set as minimizing the charging station operation and maintenance cost and user queuing time based on M / M / K queuing theory. The Pareto front is obtained using the NSGA II multi-objective algorithm. Evidence-based reasoning is used to select the final implementation scheme from the Pareto solution, and the scheme is validated and improved by substituting it into the user range anxiety level formula. Applying this technical solution can effectively reduce the cost of highway charging station layout planning and user anxiety levels.

[0185] This patent aims to explore the interaction between user behavior characteristics and charging station planning and layout, and to establish a multi-objective optimization model and solution algorithm for highway electric vehicle charging facilities that considers user range anxiety, power, and traffic systems, in order to obtain high-quality optimized configuration schemes. The dynamic traffic simulation model can accurately simulate the dynamic changes in electric vehicle driving characteristics and traffic flow based on OD matrix pairs, obtaining the traffic flow distribution at different locations and times on the highway network. The queuing theory model can calculate the charging waiting time for electric vehicle users based on the charging demand caused by traffic flow. A model considering the range anxiety level of users' psychological characteristics is established to validate and improve the layout planning results. By using the NSGA II algorithm to solve the multi-objective function and establishing an evaluation system for charging station planning through evidence reasoning, this patent aims to improve power quality, the economic operation of charging stations, the application and development of electric vehicles, promote the stable operation of the power distribution network, and provide guidance for research on the layout planning of highway electric vehicle charging facilities. This will promote the healthy and sustainable development of electric vehicles, contribute to energy conservation and emission reduction, prevent air pollution, achieve carbon neutrality, and improve social operational efficiency.

Claims

1. A highway charging facility layout planning method considering range anxiety, characterized in that, Includes the following steps: Step S1: Generate OD matrix pairs based on user driving data recorded by the highway toll system; Step S2: A method based on the OD matrix to update the traffic flow information of each road using the road model and node model in the road segment transmission model, and to reflect the user's charging demand based on the traffic flow information; Step S3: Establish candidate sites and impose distance restrictions between charging stations based on users' range anxiety and charging needs, thereby obtaining a pre-selection of candidate site schemes; Step S4: Set the objective function to minimize the charging station operation and maintenance cost and the user queuing time, and calculate the queuing time based on the M / M / K queuing theory. Step S5: Use the NSGAⅡ multi-objective algorithm to obtain the Pareto front; Step S6: Use evidence-based reasoning to select the final implementation plan from the Pareto solution, and substitute it into the user's mileage anxiety formula to verify and improve the plan; Step S3 specifically includes: Step 31: Calculate the current range anxiety level of electric vehicle users based on the number of charging stations and charging piles; （1） In the formula, Indicates electric vehicle users Given the current number of charging stations, the magnitude of range anxiety is assumed to follow a certain order. The normal distribution of , a represents the average range anxiety level of electric vehicle users, that is, the minimum SOC that electric vehicle users can tolerate when arriving at the charging station, and b is its corresponding variance. Step 32: Based on the range anxiety levels of users according to a normal distribution, for users with a confidence interval probability of 95.45% and a confidence interval probability of 99.74%, the maximum value in the set of all users' minimum SOC values ​​at arrival is used to limit the distance between charging stations. That is, 95.45% and 99.74% of electric vehicle users will have an SOC value at arrival greater than their minimum SOC value at arrival under this distance limit. Step 33: Substitute the final candidate results into the range anxiety formula to obtain the change in user range anxiety under this charging station layout scheme, i.e. （2） （3） In the formula, This represents the total number of charging stations; , These are the corresponding hyperparameters; as the number of charging piles in the charging station increases, The level of range anxiety among users given the current charging station layout; To assess users' mileage anxiety level after implementing the optimal layout plan; For electric vehicles Battery capacity; For electric vehicles The maximum battery capacity required to reach the charging station in the layout planning results; Step 34: Based on the user's range anxiety level obtained in Step 33, use the maximum value in the set of lowest SOCs with confidence interval probabilities of 95.45% and 99.74% to calculate the distance limit between charging stations until the user's range anxiety level and the distance between charging stations remain basically unchanged. （4） In the formula, Distance restrictions between charging stations; For electric vehicles The maximum battery capacity required to reach the charging station in the layout planning results; For the driving range of electric vehicles.

2. The highway charging facility layout planning method considering range anxiety according to claim 1, characterized in that, Step 2 specifically includes: Step S2 specifically includes the following steps: Step 21: Analyze the National Day data based on the gantry information of the Fujian section of G15 Shenhai Expressway to obtain the average OD matrix pair; Step 22: For each section of the highway route, specify the route length. Blocking density Free flow velocity , the maximum traffic capacity of the road Configure parameters; Step 23: Calculate the transmission capacity of each road segment separately. and receiving capabilities It is related to the cumulative traffic volume of the upstream and downstream road sections and the parameter settings in step 22; （5） （6） In the formula, , , , , , and The first The maximum capacity of the road, the location of the upstream inlet, the location of the downstream outlet, the length, the free flow velocity, the shock wave reverse propagation velocity, and the blockage density; Indicates the change over time; This indicates taking the minimum value within the parentheses; Indicates the first The upstream entrance of the road Cumulative traffic flow at any given time; Indicates the first The upstream entrance of the road Cumulative traffic flow at any given time; Step 24: Establish the node model in the dynamic road segment transmission model, substitute the cumulative traffic flow of each road segment obtained in Step 23 into the node model, and obtain the transfer flow between adjacent road segments according to different node types; Step 25: Based on the inflow, outflow, and transfer flow of traffic for each road segment obtained in Steps 23 and 24, update the cumulative inflow and outflow traffic information for each road segment. （7） （8） In the formula, Let be the cumulative traffic flow at time t downstream of road i; To get from the road Transfer to line Traffic flow during transfers; For the line Upstream Cumulative traffic flow at any given time; Suppose for any path Passing the road and road sections ,and for The downstream section, then, （9） （10） In the formula, Let i be the downstream position of road i in path p. Cumulative traffic flow at any given time; For on the road The cumulative traffic flow upstream of the location is The corresponding time; The position of path p upstream of line j Cumulative traffic flow at any given time; This is the road segment association matrix, where 1 is set when path p contains line j, and 0 otherwise. This refers to the transfer flow that moves from road i to line j at time t; Step 26: Calculate the traffic flow density at any point on any road segment based on the cumulative traffic flow information. That is, for a road segment... Assuming point On this road segment, the transmission capacity at that point is... and ability to accept They are respectively （11） （12） In the formula, , x, v, w and These represent the upstream position, downstream position, arbitrary position on the road i, free flow velocity, turbulent back propagation velocity, and blockage density, respectively. This refers to the traffic flow information at position x on the i-th road at time t. For the upstream position on the i-th road The cumulative traffic flow at any given moment.

3. The highway charging facility layout planning method considering range anxiety according to claim 1, characterized in that, The construction and operation costs of charging stations and user psychological characteristics are used as objective functions. Different weighted objective functions are adopted according to the actual problem, and the Pareto front is obtained by calculating the multi-objective function through the NSGAⅡ algorithm. Step 41: Substitute the number of charging stations to be built, the minimum number of charging piles at each charging station if a charging station is built on each line, the distance limit between charging stations, the traffic flow information at any location and time, and the candidate station sites as parameters. Step 42: Preset the NSGAⅡ parameters, with the independent variables being the location of the charging station and the corresponding number of charging piles; Step 43: Set the objective function to the construction and operation cost of charging stations. User waiting time , respectively （13） In the formula, These are the annual operation and maintenance parameters; The discount rate; The service life of the charging station; The number of charging stations to be built along the preset route; Fixed costs for building charging stations; The fixed cost of the charging station; For the first The number of charging piles built at each charging station; （14） In the formula, This indicates that the electric vehicle is at time t. Average waiting time at each charging station; This indicates the penetration rate of electric vehicles, which is the average proportion of electric vehicles traveling on highways. At time t, the electric vehicle entered the... The probability of a charging station; This represents the traffic flow of electric vehicles at time t; This is for simulating step size; Step 44: Obtain the corresponding Pareto front by changing the weights of the objective function according to actual needs.

Images