A Site Selection Method for Electric Vehicle Charging Stations Based on Random Road Networks

Abstract

This invention discloses a method for selecting electric vehicle charging station locations based on random road networks. It simulates the randomness of the original traffic network using different traffic scenarios; calculates the energy consumption of road segments under different traffic scenarios; constructs an extended network and calculates the electricity required to complete the journey using the road segment energy consumption; establishes a comprehensive location model with the goal of minimizing the construction cost of charging stations and the amount of charging during the journey under the extended network; optimizes the comprehensive location model based on the constraints of the electricity required to complete the journey and solves the problem; and determines the optimal location of charging stations in the extended network and the paths of origin-destination pairs under different traffic scenarios based on the solution results. The charging station location model of this invention can quickly and effectively select and optimize the location of charging stations. While meeting the travel needs of all origin-destination pairs, it also considers the construction cost of charging stations and the amount of charging during the journey for all origin-destination pairs, comprehensively considering road segment information under different scenarios to find the most economical electric vehicle charging station location scheme.

Description

Technical Field

[0001] This invention relates to the field of road traffic in transportation planning and management, and particularly to a method for selecting the location of electric vehicle charging stations based on random road networks. Background Technology

[0002] With the current global shortage of oil resources and increasingly serious environmental pollution, electric vehicles, as a representative of new energy vehicles, have achieved rapid development of the new energy vehicle industry and have received strong promotion from the government and widespread attention from society.

[0003] Charging infrastructure, as a crucial supporting infrastructure for ensuring long-distance travel of electric vehicles, directly impacts the operational efficiency of electric vehicles. A well-planned layout and construction of charging stations can maximize the satisfaction of electric vehicle users' needs and promote the rapid development of the new energy industry. However, the current development of electric vehicles and charging stations is uncoordinated, with problems such as low utilization rates of charging piles, slow charging speeds, and a severe shortage of charging piles in some areas becoming increasingly prominent, thus restricting the use cases of electric vehicles.

[0004] Existing charging station planning and layout methods have the following problems: (1) Most charging station layouts are based on deterministic road networks, which can only meet the charging needs of electric vehicles in a single scenario; (2) They use road segment length, road segment travel time, or road segment travel cost as road segment weights and conduct site selection research on charging stations based on service coverage for traffic travel needs. However, simply studying the site selection problem of charging stations under deterministic conditions obviously cannot meet the actual needs of traffic travel. Therefore, how to scientifically construct charging facilities and effectively improve the utilization efficiency of charging facilities is an urgent problem to be solved. Summary of the Invention

[0005] Purpose of the invention: To address the above problems, the purpose of this invention is to provide a method for selecting electric vehicle charging station locations based on random road networks. This method obtains information on different scenarios of actual traffic networks from historical data, builds a comprehensive location optimization model, meets residents' travel and charging needs, facilitates residents' lives, improves the utilization rate of charging stations while saving construction costs, and enhances the overall benefits of the electric vehicle industry.

[0006] Technical solution: The present invention provides a method for selecting the location of electric vehicle charging stations based on random road networks, comprising:

[0007] The randomness of the original traffic network is simulated using different traffic scenarios, and the probability of different traffic scenarios occurring is determined.

[0008] Based on the relationship between road segment energy consumption and road segment travel time, calculate road segment energy consumption under different traffic scenarios;

[0009] An extended network is built on the basis of the original transportation network, and the electricity required to complete the journey is calculated by utilizing the energy consumption of road segments under the extended network.

[0010] Based on randomness, a comprehensive site selection model is established with the goal of minimizing the construction cost of charging stations under the extended network and the amount of charging on the way.

[0011] The integrated location model is optimized based on the constraint of the electricity required to complete the trip. The optimized integrated location model is solved, and the optimal location of the charging station in the extended network and the origin-destination pair path under different traffic scenarios are determined based on the solution results.

[0012] Furthermore, based on the relationship between road segment energy consumption and road segment travel time, the specific calculations of road segment energy consumption under different traffic scenarios include:

[0013] Based on the relationship between vehicle speed and road segment travel time, v = l / t, and utilizing the road segment travel time under different traffic scenarios... Calculate the driving speed on the road segment under different scenarios. Where l represents the length of a road segment in the original transportation network;

[0014] Based on the extracted historical energy consumption data and corresponding speed data of the road segment, a function is used to fit the relationship between road segment energy consumption and road segment travel speed. Then, based on this relationship, road segment energy consumption under different scenarios is calculated using road segment travel speeds in different scenarios.

[0015] Furthermore, an extended network is built upon the original transportation network. Within this extended network, the electricity required to complete the journey is calculated using road segment energy consumption. Specifically, this includes:

[0016] Original transportation network G o =(N o A ° ) is the actual set of nodes N ° and the actual road segment set A ° Composition, in the actual node set N ° Each OD pair adds a set of virtual origins. and virtual endpoint From the actual set of nodes N o A virtual node set N is composed of a virtual starting point and a virtual ending point;

[0017] In the actual road segment set A o Based on this, a virtual road segment is added to each node pair (i,j) that meets the energy consumption limit. The weight of the road segment is the energy consumed by the shortest path of node pair (i,j). The virtual road segment set A is composed of the actual road segment set and the virtual road segments;

[0018] An extended network G = (N, A) is composed of a set of virtual nodes N and a set of virtual road segments A;

[0019] Using the amount of electricity consumed by the vehicle after traversing all road sections and the battery level at the destination. and starting battery power Calculations are performed to obtain the electricity required for the k-th OD pair to complete the journey in traffic scenario s. The expression is:

[0020]

[0021] In the formula, This represents the battery charge at the starting and ending points in traffic scenario s, where i represents a node, i∈∪ k∈K {O k D k}, This represents the energy consumption of a vehicle passing through road segment (i,j) in traffic scenario s. Let be the decision variable, representing whether the k-th OD pair uses road segment (i,j) in traffic scenario s. Indicates the use of road segment (i,j), when K represents the set of OD pairs and S represents the set of traffic scenarios.

[0022] Furthermore, the virtual road segment set includes A1, A2, A3, A4, and A5, and meets the following requirements:

[0023] Virtual road segment set A1 is the connection virtual starting point. From the actual starting point O k The set of road segments between them, and the energy consumption of the virtual road segment set A1 is specified. A value of 0 indicates:

[0024]

[0025] Virtual road segment set A2 connects to the actual destination D. k With virtual endpoint The set of road segments between them, and the energy consumption of the virtual road segment set A2 is specified. A value of 0 indicates:

[0026]

[0027] The electricity consumed by the shortest path between any node pair (i,j) in the original traffic network under traffic scenario s. If the energy consumption is less than the battery capacity R, then a virtual road segment (i,j,s) is created for the node pair (i,j) as the virtual road segment set A3, and the energy consumption of the virtual road segment set A3 is specified. for Represented as:

[0028]

[0029] When the virtual starting point is in traffic scenario s The amount of electricity consumed by the shortest path to any node j in the original transportation network Less than Then it is a node pair Create virtual road segments As a virtual road segment set A4, the energy consumption of the virtual road segment set A4 is specified. for Represented as:

[0030]

[0031] If, in traffic scenario s, any node i in the original traffic network is connected to the virtual endpoint... The power consumed by the shortest path Less than Then it is a node pair Create virtual road segments As a virtual road segment set A5, the energy consumption of virtual road segment set A5 is specified. for Represented as:

[0032]

[0033] Furthermore, based on randomness, a comprehensive site selection model is established with the goal of minimizing the construction cost of charging stations under the extended network and the amount of charging on the go. This model specifically includes:

[0034] Based on the probability of different traffic scenarios occurring in a random road network, a comprehensive location selection model is established to minimize the construction cost of charging stations and the amount of charging during transit. It is stipulated that all origin-destination pairs (ODs) reach their destinations under energy constraints. The expression for the comprehensive location selection model F is:

[0035]

[0036] In the formula, p s Let d represent the probability of traffic scenario s occurring. k f represents the demand flow of the k-th OD pair. i This represents the fixed cost of building a charging station at node i; Let represent the decision variable, indicating whether a charging station should be built at node i in traffic scenario s. This indicates that in traffic scenario s, a charging station needs to be built at node i. This indicates that node i does not build a charging station in scenario s.

[0037] Furthermore, the amount of electricity required to complete the trip. The constraints include start-point constraints and end-point constraints, which are expressed as follows:

[0038] When a charging station is located at the starting point, the electric vehicle's battery starts at 100% charge; when there is no charging station at the starting point, the electric vehicle's battery starts at 50% charge. The starting battery charge constraint for the electric vehicle is expressed as follows:

[0039]

[0040] When a charging station is located at the destination, the electric vehicle will be depleted of its battery upon arrival. When no charging station is located at the destination, the electric vehicle's battery level must be at least 50% before reaching the destination to ensure the continuation of the journey. The battery level constraint for the electric vehicle at the destination is expressed as follows:

[0041]

[0042] Based on the starting and ending battery power constraints, the expression for the required power to complete the journey is as follows:

[0043]

[0044] Furthermore, based on stochasticity, a comprehensive location selection model is established with the objectives of minimizing the construction cost of charging stations under the extended network and minimizing the amount of charging during the journey. This model also includes: constraining the comprehensive location selection model based on the stochastic road network while satisfying the travel demands of all origin-destination (OD) systems. These constraints include:

[0045] Charging station construction location constraints: In the extended network, it is agreed that each OD pair is limited to using nodes within the set of nodes where charging stations are being built, as follows:

[0046]

[0047] Flow balance constraint: For the k-th OD pair, a virtual starting point is defined. The outflow is 1, and the inflow is to the virtual endpoint. The flow rate is 1. In the extended network, the inflow and outflow flows of other nodes are equal, expressed as:

[0048]

[0049] Unique charging station location selection constraint: It is agreed that the construction location of charging stations is the same under different traffic scenarios, i.e., the decision variable. The value is the same across all traffic scenarios, represented as:

[0050]

[0051] Domain constraints: Define the decision variables and the battery capacity at the origin and destination points in the integrated location model. and the amount of electricity required for the trip To impose constraints, it is represented as:

[0052]

[0053]

[0054]

[0055]

[0056] Furthermore, the integrated location model is optimized based on the constraint of the electricity required to complete the trip. The optimized integrated location model is solved, and the optimal locations of charging stations in the extended network and the specific routes of origin-destination pairs under different traffic scenarios are determined based on the solution results:

[0057] The integrated location model is optimized using the constraints of battery capacity at the origin and destination points and the constraints of the integrated location model. The optimized integrated location model is then used as the electric vehicle charging station selection model for stochastic road networks, and its expression is:

[0058]

[0059]

[0060]

[0061]

[0062]

[0063]

[0064] The relevant parameters of electric vehicles, road network data, and road segment energy consumption data under different traffic scenarios are input into the electric vehicle charging station site selection model. The Gurobi solver is used to solve the model and output the optimal construction location of the charging station and the path of the OD pair under different scenarios.

[0065] Beneficial effects: Compared with the prior art, the significant advantages of this invention are:

[0066] 1. This invention creates virtual starting points and virtual ending points for each origin-end pair based on the actual road network, constructs virtual road segments for node pairs that meet energy consumption conditions, and builds an extended network; and adopts a scenario-based approach to describe the actual road network status, using historical data of road segment travel time in the road network as scenarios to represent road segment travel time, making the charging station site selection model closer to real life;

[0067] 2. Based on the relationship between driving speed and travel time, and the conversion relationship between driving speed and road segment energy consumption, a more reliable method is used to obtain road segment energy consumption, providing corresponding reference method support for road segment energy consumption in the transportation network. It is scientific, accurate, reasonable and effective.

[0068] 3. The charging station site selection model of the present invention can quickly and effectively select and optimize the location of charging stations. On the basis of meeting the travel needs of all OD pairs, it also considers the construction cost of charging stations and the on-the-go charging volume of all OD pairs. By integrating road segment information under different scenarios, it finds the most economical and applicable electric vehicle charging station site selection scheme, improves the utilization efficiency of charging stations in the real traffic network, and brings convenience to electric vehicle users. Attached Figure Description

[0069] Figure 1 A flowchart of a method for selecting electric vehicle charging station locations based on random road networks;

[0070] Figure 2 This is the original road network map in one embodiment;

[0071] Figure 3 This is a partial extended network diagram of one embodiment. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments.

[0073] Figure 1 The flowchart illustrates an electric vehicle charging station site selection method based on a random road network, as shown in one embodiment. The method includes the following steps:

[0074] (2) Use different traffic scenarios to simulate the randomness of the original traffic network and determine the probability of different traffic scenarios.

[0075] The operation of the traffic network varies at different times. Step (2) above uses multiple scenarios to characterize the actual operation of the traffic network, thereby simulating the randomness of the actual traffic network. The frequency p of the traffic scenarios occurring at different times is given respectively. s This includes weekdays, rest days, and morning and evening rush hours.

[0076] (3) Calculate the energy consumption of road segments under different traffic scenarios based on the relationship between road segment energy consumption and road segment travel time;

[0077] The energy consumption of the aforementioned road segment refers to the amount of electricity consumed by an electric vehicle when it completes a certain road segment. When the length of the road segment is determined, there is a functional relationship between the energy consumption of the road segment and the travel time of the road segment. By obtaining the travel time of the road segment in advance, the energy consumption of the road segment under different traffic scenarios can be calculated.

[0078] (4) Build an extended network on the basis of the original transportation network, and calculate the electricity required to complete the journey by utilizing the energy consumption of road segments under the extended network;

[0079] (5) Based on randomness, a comprehensive site selection model is established with the goal of minimizing the construction cost of charging stations under the extended network and the amount of charging on the way.

[0080] (6) Optimize the integrated location model based on the constraints of the electricity required to complete the trip, solve the optimized integrated location model, and determine the optimal location of the extended network charging station and the origin-destination pair path under different traffic scenarios based on the solution results.

[0081] The aforementioned method for selecting electric vehicle charging station locations based on random road networks utilizes multi-scenario simulations of actual road network traffic conditions. It extracts road segment travel times under different scenarios from the actual road network to describe the randomness of the actual traffic network, analyzes the road network status and charging demand under different scenarios, and derives an economical, reasonable, efficient, and feasible charging station construction scheme through model optimization. Simultaneously, it leverages the functional relationship between road segment travel time and energy consumption to obtain the road segment energy consumption of electric vehicles in a more reliable manner, making the location selection method more closely aligned with reality and demonstrating excellent practicality.

[0082] In one embodiment, step (2) is preceded by:

[0083] (1) Extract the following information from the city to be planned: intersection and road segment information of the actual traffic network, relevant parameters such as the battery capacity R of electric vehicles, and the site selection and construction cost f of electric vehicle charging stations. i It obtains data on residents' travel demand for electric vehicle charging stations at potential sites, as well as historical traffic data and power grid statistics.

[0084] In one embodiment, step (3) specifically includes:

[0085] Based on the relationship between vehicle speed and road segment travel time, v = l / t, and utilizing the road segment travel time under different traffic scenarios... Calculate the driving speed on the road segment under different scenarios. Where l represents the length of a road segment in the original transportation network;

[0086] Based on the extracted historical energy consumption data and corresponding speed data of the road segment, a function is used to fit the relationship between road segment energy consumption and road segment travel speed. Then, based on this relationship, road segment energy consumption under different scenarios is calculated using road segment travel speeds in different scenarios.

[0087] In one embodiment, step (4) specifically includes:

[0088] Original transportation network G o =(N o A o ) is the actual set of nodes N o and the actual road segment set A o Composition, in the actual node set N o Each OD pair adds a set of virtual origins. and virtual endpoint From the actual set of nodes N o A virtual node set N is composed of a virtual starting point and a virtual ending point;

[0089] In the actual road segment set A o Based on this, a virtual road segment is added to each node pair (i,j) that meets the energy consumption limit. The weight of the road segment is the energy consumed by the shortest path of node pair (i,j). The virtual road segment set A is composed of the actual road segment set and the virtual road segments;

[0090] An extended network G = (N, A) is composed of a set of virtual nodes N and a set of virtual road segments A;

[0091] Using the amount of electricity consumed by the vehicle after traversing all road sections and the battery level at the destination. and starting battery power Calculations are performed to obtain the electricity required for the k-th OD pair to complete the journey in traffic scenario s. The expression is:

[0092]

[0093] In the formula, This represents the battery charge at the starting and ending points in traffic scenario s, where i represents a node, i∈∪ k∈K {O k D k}, This represents the energy consumption of a vehicle passing through road segment (i,j) in traffic scenario s. Let be the decision variable, representing whether the k-th OD pair uses road segment (i,j) in traffic scenario s. Indicates the use of road segment (i,j), when K represents the set of OD pairs and S represents the set of traffic scenarios.

[0094] In one embodiment, the virtual road segment set includes A1, A2, A3, A4, and A5, satisfying the following requirements:

[0095] Virtual road segment set A1 is the connection virtual starting point. From the actual starting point O k The set of road segments between them, and the energy consumption of the virtual road segment set A1 is specified. A value of 0 indicates:

[0096]

[0097] Virtual road segment set A2 connects to the actual destination D. k With virtual endpoint The set of road segments between them, and the energy consumption of the virtual road segment set A2 is specified. A value of 0 indicates:

[0098]

[0099] The electricity consumed by the shortest path between any node pair (i,j) in the original traffic network under traffic scenario s. If the energy consumption is less than the battery capacity R, then a virtual road segment (i,j,s) is created for the node pair (i,j) as the virtual road segment set A3, and the energy consumption of the virtual road segment set A3 is specified. for Represented as:

[0100]

[0101] When the virtual starting point is in traffic scenario s The amount of electricity consumed by the shortest path to any node j in the original transportation network Less than Then it is a node pair Create virtual road segments As a virtual road segment set A4, the energy consumption of the virtual road segment set A4 is specified. for Represented as:

[0102]

[0103] If, in traffic scenario s, any node i in the original traffic network is connected to the virtual endpoint... The power consumed by the shortest path Less than Then it is a node pair Create virtual road segments As a virtual road segment set A5, the energy consumption of virtual road segment set A5 is specified. for Represented as:

[0104]

[0105] Collect the above-mentioned original transportation network G o The OD information includes demand flow and origin / destination locations. To facilitate modeling and calculation, an extended network G is constructed, consisting of a node set N and a road segment set A. Figure 3This is a schematic diagram of an extended network structure in one embodiment, where nodes 0-4 represent nodes in the original traffic network. (2,0) is used as a start-end pair for partial extension, representing a road segment. Belongs to virtual road segment set A1, road segment Segment (0, 4) belongs to virtual segment set A2 and belongs to virtual segment set A3. Belongs to virtual road segment set A4, road segment It belongs to the virtual road segment set A5.

[0106] In one embodiment, step (5) specifically includes:

[0107] Based on the probability of different traffic scenarios occurring in a random road network, a comprehensive location selection model is established to minimize the construction cost of charging stations and the amount of charging during transit. It is stipulated that all origin-destination pairs (ODs) reach their destinations under energy constraints. The expression for the comprehensive location selection model F is:

[0108]

[0109] In the formula, p s Let d represent the probability of traffic scenario s occurring. k f represents the demand flow of the k-th OD pair. i This represents the fixed cost of building a charging station at node i; Let represent the decision variable, indicating whether a charging station should be built at node i in traffic scenario s. This indicates that in traffic scenario s, a charging station needs to be built at node i. This indicates that node i does not build a charging station in scenario s.

[0110] In one embodiment, the amount of electricity required to complete the trip... The constraints include start-point constraints and end-point constraints, which are expressed as follows:

[0111] When a charging station is located at the starting point, the electric vehicle's battery starts at 100% charge; when there is no charging station at the starting point, the electric vehicle's battery starts at 50% charge. The starting battery charge constraint for the electric vehicle is expressed as follows:

[0112]

[0113] When a charging station is located at the destination, the electric vehicle will be depleted of its battery upon arrival. When no charging station is located at the destination, the electric vehicle's battery level must be at least 50% before reaching the destination to ensure the continuation of the journey. The battery level constraint for the electric vehicle at the destination is expressed as follows:

[0114]

[0115] Based on the starting and ending battery power constraints, the expression for the required power to complete the journey is as follows:

[0116]

[0117] In one embodiment, after step (5), the method further includes: constructing constraints for an integrated location model based on a stochastic road network, satisfying all OD (Original Demand) pairs' travel demands. These constraints include:

[0118] Charging station construction location constraints: In the extended network, it is agreed that each OD pair is limited to using nodes within the set of nodes where charging stations are being built, as follows:

[0119]

[0120] Flow balance constraint: For the k-th OD pair, a virtual starting point is defined. The outflow is 1, and the inflow is to the virtual endpoint. The flow rate is 1. In the extended network, the inflow and outflow flows of other nodes are equal, expressed as:

[0121]

[0122] Unique charging station location selection constraint: It is agreed that the construction location of charging stations is the same under different traffic scenarios, i.e., the decision variable. The value is the same across all traffic scenarios, represented as:

[0123]

[0124] Domain constraints: Define the decision variables and the battery capacity at the origin and destination points in the integrated location model. and the amount of electricity required for the trip To impose constraints, it is represented as:

[0125]

[0126]

[0127]

[0128]

[0129] In one embodiment, step (6) specifically includes:

[0130] The integrated location model is optimized using the constraints of battery capacity at the origin and destination points and the constraints of the integrated location model. The optimized integrated location model is then used as the electric vehicle charging station selection model for stochastic road networks, and its expression is:

[0131]

[0132]

[0133]

[0134]

[0135]

[0136]

[0137] The relevant parameters of electric vehicles, road network data, and road segment energy consumption data under different traffic scenarios obtained in step (3) are input into the electric vehicle charging station site selection model. The Gurobi solver is used to solve the model and output the optimal construction location of the charging station and the path of the OD pair under different scenarios.

[0138] In one example, consider a five-node traffic network G. o Using random road networks as the research object, this paper explains the implementation process of the above-mentioned electric vehicle charging station site selection method. Figure 2 As shown, circular markers represent different nodes, directed line segments between different nodes represent road segments in the traffic network, and labels on road segments represent the length l of that road segment. The specific steps include:

[0139] (1) Obtain the node set N from the road network o ={0,1,2,3,4} and the set of road segments A o The battery capacity of electric vehicles is specified as 580kWh, and the construction costs for each node are f0 = 2000kWh, f1 = 4000kWh, f2 = 6000kWh, f3 = 4000kWh, and f4 = 9000kWh. The travel demand data for OD (Original Demand) is shown in Table 1. k and D k This represents the origin and destination of the k-th OD pair.

[0140] Table 1 Travel Demand Table

[0141]

[0142] (2) Simulate the randomness of the actual traffic network using a scenario-based approach, given the probability p of different traffic scenarios occurring. s Travel time between different road sections under different traffic scenarios The tables are different; both the horizontal and vertical columns represent nodes. The road segment travel time data under different scenarios is shown in Table 2.

[0143] Table 2. Road segment travel time data under different scenarios (unit: h)

[0144]

[0145] (3) Using a set of road segment energy consumption and corresponding speed data obtained from the actual road network, as shown in Table 3, a commonly used function is selected to fit the relationship between the two, and the relationship between energy consumption and vehicle speed is obtained as follows:

[0146] Table 3. Energy Consumption and Speed ​​of Road Sections

[0147]

[0148] (4) Based on the relationship between energy consumption and road segment travel time, obtain the road segment energy consumption under different scenarios. As shown in Table 4.

[0149] Table 4. Road section energy consumption under different scenarios

[0150]

[0151] (5) In the original transportation network G ° =(N ° A ° Based on the above, an extended network G = (N, A) is built, creating virtual origin and destination points for each origin-destination (OD) pair, and creating virtual road segments for node pairs that meet energy consumption conditions. The road segment set of the extended road network is shown in Table 5.

[0152] Table 5 Energy Consumption of Road Sections in the Expanded Network

[0153]

[0154]

[0155] (6) Based on meeting all travel needs, and taking into account both the construction cost of charging stations and the travel costs of residents, construct the objective function of the charging station site selection model:

[0156]

[0157] The constraints of the model include:

[0158]

[0159]

[0160]

[0161]

[0162]

[0163]

[0164]

[0165] r i ≥0, i∈{O k ∪D k},k∈K (8)

[0166] x ijk ∈{0,1},(i,j)∈A,k∈K (9)

[0167] y i ∈{0,1},i∈N 0 (10)

[0168] Equation (1) represents the total energy required for the k-th OD to complete its journey. Equation (2) represents the battery level constraint of the starting point of OD to k: when a charging station is built, the battery level of the starting point is 100%, otherwise it is 50%. Equation (3) represents the battery level constraint of the ending point of OD to k: when a charging station is built, the battery level of the ending point is 0%, otherwise it is 50%. Equation (4) represents the flow balance constraint, which ensures that the selected road segments in each scenario form a complete path between ODs, and the inflow and outflow of other nodes in the network are equal except for the starting and ending points of the journey. Equation (5) is the construction constraint of the charging station, which means that the charging station is built at the end of the road segment used by the OD. Equation (6) is the unique charging station location constraint. In this embodiment, it is necessary to find the best charging station construction scheme. Therefore, the same charging station location layout should be selected in each scenario. If a charging station is built at node j in traffic scenario s, then this node must be selected in other scenarios. Equations (7)-(10) determine the domain of the decision variables.

[0169] (8) The optimal path of OD pair and the construction scheme of charging station in each scenario are calculated using the Gurobi solver, as shown in Table 6-7 below.

[0170] Table 6 Construction plan for establishing charging stations at each node

[0171]

[0172] (A value of 0 in the table above means that the node does not build a charging station, and a value of 1 means that the node builds a charging station.)

[0173] Table 7 Optimal Paths for OD Pairs in Various Scenarios

[0174]

[0175] The results from the model above show that the charging station location scheme is the same in all scenarios, but the optimal path for the OD pair may differ in different scenarios. For example, the optimal path for the OD pair (2,0) in scenarios 0 and 1 is 2→1→3→0, but due to the difference in energy consumption of road segments in different scenarios, the optimal path in scenario 2 becomes 2→1→4→3→0. The results of the model examples are more in line with the actual situation.

[0176] This invention, when selecting locations for electric vehicle charging stations, constructs an extended network based on the randomness of the road network. It utilizes the functional relationship between energy consumption and travel time to obtain the energy consumption of road segments under different scenarios. Then, aiming to minimize the construction cost of charging stations and the on-the-go charging volume of OD pairs, the model is simplified based on the charging behavior of electric vehicles to ultimately determine the charging station locations. Based on the random road network, this ensures that user travel is more closely aligned with actual conditions. Different travel routes will be selected in different scenarios, and the functional relationship between energy consumption and speed reflects the actual charging needs of electric vehicles. While meeting user charging needs, this invention minimizes the total cost of charging station construction and user travel, improves the utilization rate of charging stations, and facilitates residents' lives.

Claims

1. A method for selecting the location of electric vehicle charging stations based on random road networks, characterized in that, include: The randomness of the original traffic network is simulated using different traffic scenarios, and the probability of different traffic scenarios occurring is determined. Based on the relationship between road segment energy consumption and road segment travel time, calculate road segment energy consumption under different traffic scenarios; Based on the extracted historical energy consumption data and corresponding speed data of the road segment, a function is used to fit the relationship between road segment energy consumption and road segment travel speed. Then, based on this relationship, road segment energy consumption under different scenarios is calculated using road segment travel speeds in different scenarios. ; An extended network is built on the basis of the original transportation network, and the electricity required to complete the journey is calculated by utilizing the energy consumption of road segments under the extended network. Based on randomness, a comprehensive site selection model is established with the goal of minimizing the construction cost of charging stations under the extended network and the amount of charging on the way. The integrated location model is optimized based on the constraints of the electricity required to complete the trip. The optimized integrated location model is solved, and the optimal location of the charging stations in the extended network and the origin-destination pair paths under different traffic scenarios are determined based on the solution results. An extended network is built upon the original transportation network. The electricity required to complete the journey is calculated using road segment energy consumption within this extended network. Specifically, this includes: primitive transportation network From the actual set of nodes Collection of actual road sections Composition, in the actual set of nodes Each OD pair adds a set of virtual origins. and virtual endpoint , consisting of the actual set of nodes A virtual node set is composed of a virtual starting point and a virtual ending point. ; Collection on actual road sections Based on this, for each node pair that meets the energy consumption limit Add a virtual road segment with a weight equal to the number of node pairs. Energy consumed by the shortest path A virtual road segment set A is composed of a set of actual road segments and virtual road segments; The extended network consists of a set of virtual nodes N and a set of virtual road segments A. ; Using the amount of electricity consumed by the vehicle after traversing all road sections and the battery level at the destination. and starting battery power Calculations are performed to obtain the electricity required for the k-th OD pair to complete the journey in traffic scenario s. The expression is: ； In the formula, This represents the battery level at the destination in traffic scenario s. This represents the battery charge at the starting point in traffic scenario s. In traffic scenarios Exit vehicle passage section Energy consumption of road sections; Let be the decision variable, representing the situation in the traffic scenario. Next indivual Whether to use the road section ,when Indicates the road section used ,when This indicates that the road section is not in use. K represents the set of OD pairs. Represents a set of traffic scenarios; Based on randomness, a comprehensive site selection model is established with the goal of minimizing the construction cost of charging stations under the extended network and the amount of charging on the go. Specifically, it includes: Based on the probability of different traffic scenarios occurring in a random road network, a comprehensive location selection model is established to minimize the construction cost of charging stations and the amount of charging during transit. It is stipulated that all origin-destination pairs (ODs) reach their destinations under energy constraints. The expression for the comprehensive location selection model F is: ； In the formula, Representing traffic scenarios The probability of occurrence This represents the demand flow of the k-th OD pair. Indicates at node Fixed costs of building charging stations; Let represent the decision variable, and let represent the node in traffic scenario s. Whether to build charging stations, when , indicating a traffic scene The following needs to be done at the node Building charging stations, when , indicating a traffic scene Next node No charging stations will be built.

2. The method for selecting the location of electric vehicle charging stations according to claim 1, characterized in that, Based on the relationship between road segment energy consumption and road segment travel time, the calculation of road segment energy consumption under different traffic scenarios specifically includes: Based on the relationship between vehicle speed and road segment travel time Utilizing the travel time of different traffic scenarios Calculate the driving speed on the road segment under different scenarios. ;in This indicates the length of a road segment in the original transportation network.

3. The method for selecting the location of electric vehicle charging stations according to claim 1, characterized in that, The virtual road segment set includes , It must meet the following requirements: Virtual road segment set To connect virtual starting point and the actual starting point The set of road segments between them, and the definition of a virtual road segment set. Energy consumption A value of 0 indicates: ； Virtual road segment set To connect the actual endpoint With virtual endpoint The set of road segments between them, and the definition of a virtual road segment set. Energy consumption A value of 0 indicates: ； Traffic scenarios Any node pair in the original transportation network The power consumed by the shortest path Smaller than battery capacity Then it is a node pair Create virtual road segments As a set of virtual road segments And specify the virtual road segment set Consume energy for , is represented as: ； Traffic scenarios Next virtual starting point To any node in the original transportation network The power consumed by the shortest path Less than Then it is a node pair Create virtual road segments As a set of virtual road segments And specify the virtual road segment set for , is represented as: ； If traffic scenario Any node in the original transportation network To the virtual endpoint The power consumed by the shortest path Less than Then it is a node pair Create virtual road segments As a set of virtual road segments And specify the virtual road segment set Consume energy for , represented as: 。 4. The method for selecting the location of electric vehicle charging stations according to claim 1, characterized in that, Electricity required to complete the trip The constraints include start-point constraints and end-point constraints, which are expressed as follows: When a charging station is located at the starting point, the electric vehicle's battery level starts at 100%; when there is no charging station at the starting point, the electric vehicle's battery level starts at 50%. The starting battery level constraint for the electric vehicle is expressed as follows: ； When a charging station is located at the destination, the electric vehicle will be depleted of its battery upon arrival. When no charging station is located at the destination, the electric vehicle's battery level must be at least 50% before reaching the destination to ensure the continuation of the journey. The battery level constraint for the electric vehicle at the destination is expressed as follows: ； Based on the starting and ending battery power constraints, the expression for the required power to complete the journey is as follows:

5. The method for selecting the location of electric vehicle charging stations according to claim 4, characterized in that, Based on randomness, a comprehensive site selection model is established with the objectives of minimizing the construction cost of charging stations under the extended network and minimizing the amount of charging on the go. This model also includes: Under the premise of satisfying all OD (Original Demand) travel demands, the constraints for constructing an integrated location model based on a stochastic road network are as follows: Charging station construction location constraints: In the extended network, it is agreed that each OD pair is limited to using nodes within the set of nodes where charging stations are being built, as follows: ； Flow balance constraint: for the first indivual Yes, agree on a virtual starting point. The outflow is 1, and the inflow is to the virtual endpoint. The flow rate is 1. In the extended network, the inflow and outflow flows of other nodes are equal, expressed as: ； Unique charging station location selection constraint: It is agreed that the construction location of charging stations is the same under different traffic scenarios, i.e., the decision variable. The value is the same across all traffic scenarios, represented as: ； Domain constraints: Define the decision variables and the battery capacity at the origin and destination points in the integrated location model. and the amount of electricity required for the trip To impose constraints, it is represented as: ； ； ； 。 6. The method for selecting the location of an electric vehicle charging station according to claim 5, characterized in that, The integrated location model is optimized based on the constraint of the electricity required to complete the trip. The optimized integrated location model is solved, and the optimal locations of charging stations in the extended network and the specific routes of origin-destination pairs under different traffic scenarios are determined based on the solution results: The integrated location model is optimized using the constraints of battery capacity at the origin and destination points and the constraints of the integrated location model. The optimized integrated location model is then used as the electric vehicle charging station selection model for stochastic road networks, and its expression is: ； ； ； ； ； ； The relevant parameters of electric vehicles, road network data, and road segment energy consumption data under different traffic scenarios are input into the electric vehicle charging station site selection model, and then... The solver solves the problem and outputs the optimal construction location of charging stations and the results for different scenarios. The correct path.

Images