Electric vehicle charging demand prediction method and system for distinguishing vehicle types

Abstract

The invention provides an electric vehicle charging demand prediction method for distinguishing vehicle types, and the method comprises the steps: firstly considering different types of electric vehicles, determining a travel starting point and a travel ending point of the electric vehicles through a transfer matrix, and determining an optimal driving path of the electric vehicles based on the travel starting point, the travel ending point and a road weight; and then based on the optimal driving path, calculating the residual SOC of the electric vehicle at each node by adopting the established energy consumption model, judging whether the electric vehicle has a charging demand or not at each node according to the residual SOC, if so, updating the charging amount of the node, and finally generating a space-time distribution result of the charging demand of the electric vehicle. According to the method, the prediction accuracy is improved, and meanwhile, the method is more practical and operable.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric vehicle charging, and in particular relates to a method and system for predicting electric vehicle charging demand by differentiating vehicle types. Background Art

[0002] As the number of electric vehicles increases, the planning and layout of charging facilities has become a key issue. Different types of electric vehicles have different charging frequencies, charging times, and charging power requirements. If this is ignored and unified planning is carried out, it may lead to too dense charging stations in some areas, while there is a shortage of charging facilities in other areas. Therefore, it is very important to reasonably plan the construction of charging facilities according to the characteristics of different types of electric vehicles through accurate charging demand forecasts, avoid waste of resources, and ensure the efficient operation of the charging network. Summary of the invention

[0003] The purpose of the present invention is to provide a method and system for predicting charging demand of electric vehicles by differentiating vehicle types in order to solve the above problems in the prior art.

[0004] To achieve the above objectives, the technical solution of the present invention is as follows:

[0005] In a first aspect, the present invention proposes a method for predicting charging demand of electric vehicles by differentiating vehicle types, comprising:

[0006] S1, determine the starting point and end point of the electric vehicle through the transfer matrix;

[0007] S2, determining the optimal driving path of the electric vehicle based on the travel starting point, end point and road weight;

[0008] S3, considering different types of electric vehicles, based on the optimal driving path, the established energy consumption model is used to calculate the remaining SOC of the electric vehicle at each node, and at each node, it is determined whether the electric vehicle has a charging demand based on the remaining SOC. If so, the charging amount of the node is updated and then enter S4; if not, directly enter S4;

[0009] S4, determine whether the trip end point has been reached. If not, update the remaining SOC, the start time and location of the next trip, and then return to S1 for the next cycle; if reached, enter S5;

[0010] S5. Determine whether all electric vehicles have completed their journeys. If so, generate the spatiotemporal distribution results of the electric vehicle charging demand. If not, return to S1 to calculate the journey of the next electric vehicle until the charging demand prediction for all electric vehicles is completed.

[0011] In S2, the road weight is calculated according to the following formula:

[0012]

[0013] In the above formula, is the road weight of road ij in time period t, ρ1 and ρ2 are the weight factors of road impedance and road section distance respectively, d ij is the length of road ij, R ij (t) is the road impedance of road ij in period t, t0 is the travel time under zero flow conditions, l and ε are impedance influencing factors, is the saturation of road ij in period t, are the traffic volume and road capacity of road ij in period t respectively.

[0014] The energy consumption model includes:

[0015]

[0016] In the above formula, is the power required by the tires of an electric vehicle when it travels on road ij, c r , c1, c2 are rolling resistance coefficients, m and g are vehicle mass and gravity acceleration respectively. is the average speed of road ij in period t, ρ a is the air density, A f is the front area of ​​the vehicle, C D is the aerodynamic drag coefficient of the vehicle, η M , η D They are the efficiency of the vehicle motor and the transmission system, ΔO ij is the driving loss of road ij, is the distance of the k-th journey, C B , η B are battery capacity, battery efficiency, and O I , O F are the starting SOC and the remaining SOC, η N is the number of network nodes, is the intermediate parameter;

[0017] For the starting time, the starting time of buses is set to 07:30; the starting time of private cars and taxis is determined according to the following probability density function:

[0018]

[0019] In the above formula, t s is the starting travel time, α is the scale parameter, C and K are shape parameters;

[0020] For parking time, the parking time after a single trip chain for buses is set at 30 minutes; the parking time for private cars and taxis varies according to the location;

[0021] For the starting SOC, the starting SOC of the bus is set to 1; the starting SOC of private cars and taxis is determined according to the following probability density function:

[0022]

[0023] In the above formula, O I is the starting SOC, σ SOC , μ SOC are the mean and standard deviation of the Gaussian distribution.

[0024] In S1, the transfer matrix is:

[0025] G={N,R,T,W,D,Q,S,V}

[0026] N = {o i |i=1,2,…,n N}

[0027] R={(o i , o j )|i∈N,j∈N,i≠j}

[0028] T={t|t=1,2,…,24}

[0029]

[0030] D={d ij |(o i , o j )∈R}

[0031]

[0032] In the above formula, G is the set of traffic network models, N is the set of intersection nodes in the network, R is the set of roads in the network, T is the set of time periods, W is the set of road weights in the network, D is the set of actual distances between nodes, Q is the set of road traffic volume in the network, S is the set of road saturation, V is the set of road speeds, and n is the set of road traffic volume. N is the number of network nodes, o i , o j are the i-th and j-th nodes respectively, t is the time period number, is the road weight of road ij in period t, d ij is the length of road ij, are the traffic volume, saturation and average speed of road ij in period t respectively.

[0033] The S2 uses the Dijkstra algorithm to determine the optimal driving path of the electric vehicle.

[0034] In a second aspect, the present invention proposes an electric vehicle charging demand prediction system for distinguishing vehicle types, including a travel start and end point determination module, an optimal driving path calculation module, a remaining SOC calculation module, a first judgment module, a second judgment module, and a third judgment module;

[0035] The travel start point and end point determination module is used to determine the travel start point and end point of the electric vehicle through a transfer matrix;

[0036] The optimal driving path calculation module is used to determine the optimal driving path of the electric vehicle based on the travel starting point, the end point and the road weight;

[0037] The remaining SOC calculation module is used to consider different types of electric vehicles and calculate the remaining SOC of the electric vehicle at each node based on the optimal driving path using an established energy consumption model;

[0038] The first judgment module is used to judge whether the electric vehicle has a charging demand according to the remaining SOC at each node. If so, the second judgment module is started after the charging amount of the node is updated; if not, the second judgment module is directly started;

[0039] The second judgment module is used to judge whether the trip destination has been reached. If not, the remaining SOC, the start time and location of the next trip are updated and sent to the travel start and end point determination module for the next cycle; if it has been reached, the third judgment module is started;

[0040] The third judgment module is used to judge whether all electric vehicles have completed their journeys. If so, the spatiotemporal distribution results of the electric vehicle charging requirements are generated. If not, the travel start and end point determination module is started to calculate the journey of the next electric vehicle until the charging requirements of all electric vehicles are predicted.

[0041] The road weight is calculated according to the following formula:

[0042]

[0043] In the above formula, is the road weight of road ij in time period t, ρ1 and ρ2 are the weight factors of road impedance and road section distance respectively, d ij is the length of road ij, R ij (t) is the road impedance of road ij in period t, t0 is the travel time under zero flow conditions, l and ε are impedance influencing factors, is the saturation of road ij in period t, are the traffic volume and road capacity of road ij in period t respectively.

[0044] The energy consumption model includes:

[0045]

[0046] In the above formula, is the power required by the tires of an electric vehicle when it travels on road ij, c r , c1, c2 are rolling resistance coefficients, m and g are vehicle mass and gravity acceleration respectively. is the average speed of road ij in period t, ρ a is the air density, A f is the front area of ​​the vehicle, C D is the aerodynamic drag coefficient of the vehicle, η M , η D They are the efficiency of the vehicle motor and the transmission system, ΔO ij is the driving loss of road ij, is the distance of the kth journey, C B , η B are battery capacity, battery efficiency, and O I , O F are the starting SOC and the remaining SOC, n N is the number of network nodes, is the intermediate parameter;

[0047] For the starting time, the starting time of buses is set to 07:30; the starting time of private cars and taxis is determined according to the following probability density function:

[0048]

[0049] In the above formula, t s is the starting travel time, α is the scale parameter, C and K are shape parameters;

[0050] For parking time, the parking time after a single trip chain for buses is set at 30 minutes; the parking time for private cars and taxis varies according to the location;

[0051] For the starting SOC, the starting SOC of the bus is set to 1; the starting SOC of private cars and taxis is determined according to the following probability density function:

[0052]

[0053] In the above formula, O I is the starting SOC, μ SOC , μ SOCare the mean and standard deviation of the Gaussian distribution.

[0054] The transfer matrix is:

[0055] G={N,R,T,W,D,Q,S,V}

[0056] N = {o i |i=1,2,…,n N}

[0057] R={(o i , o j )|∈N,j∈N,i≠j}

[0058] T={t|t=1,2,…,24}

[0059]

[0060] D={d ij |(o i , o j )∈R}

[0061]

[0062] In the above formula, G is the set of traffic network models, N is the set of intersection nodes in the network, R is the set of roads in the network, T is the set of time periods, W is the set of road weights in the network, D is the set of actual distances between nodes, Q is the set of road traffic volume in the network, S is the set of road saturation, V is the set of road speeds, and n is the set of road traffic volume. N is the number of network nodes, o i , o j are the i-th and j-th nodes respectively, t is the time period number, is the road weight of road ij in period t, d ij is the length of road ij, are the traffic volume, saturation and average speed of road ij in period t respectively.

[0063] The optimal driving path calculation module uses the Dijkstra algorithm to determine the optimal driving path of the electric vehicle.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] 1. The present invention proposes a method for predicting the charging demand of electric vehicles by distinguishing vehicle types. The method first considers different types of electric vehicles, determines the starting point and end point of the electric vehicle through the transfer matrix, and determines the optimal driving path of the electric vehicle based on the starting point, end point and road weight. Then, based on the optimal driving path, the established energy consumption model is used to calculate the remaining SOC of the electric vehicle at each node, and at each node, it is determined whether the electric vehicle has a charging demand based on the remaining SOC. If so, the charging amount of the node is updated, and finally the spatiotemporal distribution result of the charging demand of the electric vehicle is generated. On the one hand, the method takes into account the differences in charging demand, endurance and charging habits of different types of EVs. By analyzing these differences in a targeted manner, it can provide a more personalized charging demand prediction and improve the accuracy of the prediction; on the other hand, the method not only considers the number and type of electric vehicles, but also factors such as road weight and driving power consumption are included in the calculation, making the charging demand prediction more realistic and operational.

[0066] 2. In a method for predicting the charging demand of electric vehicles that distinguishes vehicle types proposed in the present invention, the energy consumption model comprehensively considers factors such as mass, rolling resistance coefficient, aerodynamics, and drag coefficient, dynamically adapts to different road conditions and driving conditions, and carefully depicts the energy consumption composition of the vehicle during driving; the transfer matrix considers real-time traffic factors including traffic flow, saturation, speed, etc. These factors can accurately reflect the real-time traffic conditions. Based on the transfer matrix, optimal path planning, and energy consumption model, the charging demand of electric vehicles can be predicted more finely, the flexibility and response speed of the charging network can be improved, and the rough estimation problems that may occur in traditional methods can be avoided. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 This is a flow chart of the method described in Example 1.

[0068] Figure 2 This is a schematic diagram of the structure of the system described in Example 2. DETAILED DESCRIPTION

[0069] The present invention is further described in detail below in conjunction with specific implementations and drawings.

[0070] The present invention proposes a method for predicting the charging demand of electric vehicles by distinguishing vehicle types. The method not only considers the number and type of electric vehicles, but also factors such as road weight and driving power consumption are included in the calculation, making the charging demand prediction more realistic and operational. By updating the remaining power (SOC) and charging demand of the vehicle in real time, the model can adapt to changes in different travel situations and improve the flexibility and response speed of the electric vehicle charging network. In addition, the method also considers and analyzes the differences in charging demand, endurance and charging habits among different types of EVs. By analyzing these differences in a targeted manner, it can provide more personalized charging demand predictions and improve the accuracy of the predictions. The present invention helps to solve the problem of dynamic prediction of the distribution of electric vehicle charging demand, and provides a scientific basis for the site selection and planning of charging stations, load management of power systems and orderly control of electric vehicle charging behavior.

[0071] Embodiment 1:

[0072] A method for predicting electric vehicle charging demand by distinguishing vehicle types, such as Figure 1 As shown, the specific implementation steps are as follows:

[0073] 1. Determine the parameters of the model, including the total number of electric vehicles N max , type T, starting SOC and other parameter data.

[0074] 2. Use graph theory analysis to establish a dynamic road network topology and use the transfer matrix to determine the starting and ending points of each electric vehicle.

[0075] The transfer matrix is:

[0076] G={N,R,T,W,D,Q,S,V}

[0077] N = {o i |i=1,2,…,n N}

[0078] R={(o i , o j )|i∈N,j∈N,i≠j}

[0079] T={t|t=1,2,…,24}

[0080]

[0081] D={d ij |(o i , o j )∈R}

[0082]

[0083] In the above formula, G is the set of traffic network models, that is, the overall structure of the entire traffic network; N is the set of intersection nodes in the network, representing the starting and ending points of all trips. Nodes can be road intersections, corners or points of interest (such as charging stations, etc.); R is the set of roads in the network, representing the road paths between nodes. Each road section connects two nodes to form a driving path; T is the time period set, dividing a day into 24 hours, that is, T = {1, 2, 3, ..., 24}; W is the road weight set in the network, which is used to calculate the optimal path; D is the actual distance set between each node, indicating that each pair of nodes o i and j The length of the road between two points, rather than the straight-line distance between two points; Q is the road traffic volume set of the road network; S is the road saturation set, which is used to describe the congestion of the road. A value close to 1 indicates that the road section is close to full load; V is the road speed set, which is used to estimate the travel time of vehicles in different time periods, n N is the number of road network nodes, indicating the number of all intersection nodes included in the model, o i , o j are the i-th and j-th nodes respectively, t is the time period number, is the road weight of road ij in period t, d ij is the length of road ij, are the traffic volume, saturation and average speed of road ij in period t respectively.

[0084] 3. Based on the actual road traffic conditions, a road resistance model was established to determine the road weight.

[0085] The road resistance model includes:

[0086]

[0087] In the above formula, R ij (t) is the road impedance of road ij at time period t. When it is smaller, the road impedance is relatively low and the travel speed is faster; When it is close to 1 (close to full load), the impedance increases significantly, which means that serious congestion may occur. t0 is the travel time under zero flow conditions, l and ε are impedance influencing factors, which can be obtained by fitting actual traffic data (such as travel time and saturation under different flow rates). is the saturation of road ij in period t, are the traffic volume and road capacity of road ij in period t respectively. Different road grades have different capacities.

[0088] The road weight is calculated according to the following formula:

[0089]

[0090] In the above formula, is the road weight of road ij in time period t, ρ1 and ρ2 are the weight factors of road impedance and road section distance respectively, d ij is the length of road ij.

[0091] 4. Based on the starting point, end point and road weight of the trip, the Dijkstra algorithm is used to determine the optimal driving path of the electric vehicle. In the Dijkstra algorithm, the road weight data is used as the path weight.

[0092] 5. Considering different types of electric vehicles, the established energy consumption model is used to calculate the remaining SOC of electric vehicles at each node based on the optimal driving path.

[0093] This embodiment divides electric vehicles into three categories: private cars, taxis and buses, and differentiates them in three key parameters, specifically:

[0094] Starting time: The starting time of buses is relatively fixed, and 07:30 is taken as the starting time; the starting time of private cars and taxis is relatively random, and it is assumed that they follow the Burr XII distribution. Its probability density function is:

[0095]

[0096] In the above formula, t s is the starting travel time, α is the scale parameter, which determines the time scale of the starting travel time, C and K are shape parameters, C reflects the distribution intensity of travel time in the main concentrated period, and K affects the distribution range of travel time.

[0097] In this embodiment, α=7.896, C=6.696, and K=0.609.

[0098] Parking time: Taking into account the operation mode of buses, the parking time after a single trip chain is set at 30 minutes; the parking time for private cars and taxis varies according to different locations.

[0099] Starting SOC (battery state of charge): For buses, the starting SOC is assumed to be 1 (i.e. fully charged); for private cars and taxis, the starting SOC follows a Gaussian distribution, and its probability density function is:

[0100]

[0101] In the above formula, O I is the starting SOC, σ SOC , μ SOC are the mean and standard deviation of the Gaussian distribution.

[0102] The above differentiated parameter settings enable different types of electric vehicles to exhibit unique travel and charging behaviors in the model, which is more in line with actual traffic and charging needs.

[0103] Energy consumption models include:

[0104]

[0105] In the above formula, is the power required by the tires of an electric vehicle when it travels on road ij, c r , c1, c2 are rolling resistance coefficients, c r =1.75, c1=0.0328, c2=4.575, m and g are the vehicle mass and gravitational acceleration respectively, is the average speed of road ij in period t, ρ a is the air density, A f is the front area of ​​the vehicle, C D is the aerodynamic drag coefficient of the vehicle, η M , η D They are the efficiency of the vehicle motor and the transmission system, ΔO ij is the driving loss of road ij. According to the characteristics of different road sections, such as slope, road conditions, etc., the model will calculate the corresponding loss ΔO ij , is the distance of the k-th journey, C B , η B are battery capacity, battery efficiency, and O I , O F are the starting SOC and the remaining SOC, n N is the number of network nodes, is the intermediate parameter.

[0106] The model predicts the energy consumption of electric vehicles based on different road sections and driving conditions, and calculates their power requirements under different driving scenarios.

[0107] 6. At each node, determine whether the electric vehicle has a charging demand based on the remaining SOC. If so, update the node's charging capacity (i.e., the energy consumption when the node meets the charging demand) and then proceed to step 7; if not, directly proceed to step 7.

[0108] 7. Determine whether the end of the trip has been reached. If not, update the remaining SOC, the start time and location of the next trip, and then return to step 1 for the next cycle; if it has been reached, proceed to step 8;

[0109] 8. Determine whether all electric vehicles have completed their journeys. If so, generate the spatiotemporal distribution results of the electric vehicle charging demand. If not, return to step 1 to calculate the journey of the next electric vehicle until the charging demand prediction for all electric vehicles is completed.

[0110] Embodiment 2:

[0111] An electric vehicle charging demand prediction system that distinguishes vehicle types, such as Figure 2 As shown, it includes a travel starting point and end point determination module, an optimal driving path calculation module, a remaining SOC calculation module, a first judgment module, a second judgment module, and a third judgment module.

[0112] The travel start point and end point determination module is used to determine the travel start point and end point of the electric vehicle through a transfer matrix, and the transfer matrix is:

[0113] G={N,R,T,W,D,Q,S,V}

[0114] N = {o i |i=1,2,…,n N}

[0115] R={(o i , o j )|i∈N,j∈N,i≠j}

[0116] T={t|t=1,2,…,24}

[0117]

[0118] D={d ij |(o i , o j )∈R}

[0119]

[0120] In the above formula, G is the set of traffic network models, N is the set of intersection nodes in the network, R is the set of roads in the network, T is the set of time periods, W is the set of road weights in the network, D is the set of actual distances between nodes, Q is the set of road traffic volume in the network, S is the set of road saturation, V is the set of road speeds, and n is the set of road traffic volume. N is the number of network nodes, o i , o j are the i-th and j-th nodes respectively, t is the time period number, is the road weight of road ij in period t, d ij is the length of road ij, are the traffic volume, saturation and average speed of road ij in period t respectively.

[0121] The optimal driving path calculation module is used to determine the optimal driving path of the electric vehicle using the Dijkstra algorithm based on the travel starting point, the end point and the road weight, wherein the road weight is calculated according to the following formula:

[0122]

[0123]

[0124] In the above formula, is the road weight of road ij in time period t, ρ1 and ρ2 are the weight factors of road impedance and road section distance respectively, d ij is the length of road ij, R ij (t) is the road impedance of road ij in period t, t0 is the travel time under zero flow conditions, l and v are impedance influencing factors, is the saturation of road ij in period t, are the traffic volume and road capacity of road ij in period t respectively;

[0125] In the Dijkstra algorithm, the road weight data is used as the path weight.

[0126] The remaining SOC calculation module is used to consider different types of electric vehicles, based on the optimal driving path, and use the established energy consumption model to calculate the remaining SOC of the electric vehicle at each node, wherein the energy consumption model includes:

[0127]

[0128] In the above formula, is the power required by the tires of an electric vehicle when it travels on road ij, c r , c1, c2 are rolling resistance coefficients, m and g are vehicle mass and gravity acceleration respectively. is the average speed of road ij in period t, ρ a is the air density, A f is the front area of ​​the vehicle, C D is the aerodynamic drag coefficient of the vehicle, η M , η D They are the efficiency of the vehicle motor and the transmission system, ΔO ij is the driving loss of road ij, is the distance of the k-th journey, C B , η B are battery capacity, battery efficiency, and O I , O F are the starting SOC and the remaining SOC, n N is the number of network nodes, is the intermediate parameter;

[0129] For the starting time, the starting time of buses is set to 07:30; the starting time of private cars and taxis is determined according to the following probability density function:

[0130]

[0131] In the above formula, t s is the starting travel time, α is the scale parameter, C and K are shape parameters;

[0132] For parking time, the parking time after a single trip chain for buses is set at 30 minutes; the parking time for private cars and taxis varies according to the location;

[0133] For the starting SOC, the starting SOC of the bus is set to 1; the starting SOC of private cars and taxis is determined according to the following probability density function:

[0134]

[0135] In the above formula, O I is the starting SOC, σ SOC , μ SOC are the mean and standard deviation of the Gaussian distribution.

[0136] The first judgment module is used to judge whether the electric vehicle has a charging demand according to the remaining SOC at each node. If so, the second judgment module is started after the charging amount of the node is updated; if not, the second judgment module is directly started;

[0137] The second judgment module is used to judge whether the trip destination has been reached. If not, the remaining SOC, the start time and location of the next trip are updated and sent to the travel start and end point determination module for the next cycle; if it has been reached, the third judgment module is started;

[0138] The third judgment module is used to judge whether all electric vehicles have completed their journeys. If so, the spatiotemporal distribution results of the electric vehicle charging requirements are generated. If not, the travel start and end point determination module is started to calculate the journey of the next electric vehicle until the charging requirements of all electric vehicles are predicted.

Claims

1. A method for predicting charging demand of electric vehicles by distinguishing vehicle types, characterized in that: The method comprises: S1, determine the starting point and end point of the electric vehicle through the transfer matrix; S2, determining the optimal driving path of the electric vehicle based on the travel starting point, end point and road weight; S3, considering different types of electric vehicles, based on the optimal driving path, the established energy consumption model is used to calculate the remaining SOC of the electric vehicle at each node, and at each node, it is determined whether the electric vehicle has a charging demand based on the remaining SOC. If so, the charging amount of the node is updated and then enter S4; if not, directly enter S4; S4, determine whether the end of the trip has been reached. If not, update the remaining SOC, the start time and location of the next trip, and then return to S1 for the next cycle; if it has been reached, enter S5; S5. Determine whether all electric vehicles have completed their journeys. If so, generate the spatiotemporal distribution results of the electric vehicle charging demand. If not, return to S1 to calculate the journey of the next electric vehicle until the charging demand prediction for all electric vehicles is completed.

2. The method for predicting charging demand of electric vehicles by distinguishing vehicle types according to claim 1, characterized in that: In S2, the road weight is calculated according to the following formula: In the above formula, is the road weight of road ij in time period t, ρ1 and ρ2 are the weight factors of road impedance and road section distance respectively, d ij is the length of road ij, R ij (t) is the road impedance of road ij in period t, t0 is the travel time under zero flow conditions, l and ε are impedance influencing factors, is the saturation of road ij in period t, are the traffic volume and road capacity of road ij in period t respectively.

3. A method for predicting charging demand of electric vehicles by distinguishing vehicle types according to claim 1 or 2, characterized in that: The energy consumption model includes: In the above formula, is the power required by the tires of an electric vehicle when it travels on road ij, c r , c1, c2 are rolling resistance coefficients, m and g are vehicle mass and gravity acceleration respectively. is the average speed of road ij in period t, ρ a is the air density, A f is the front area of ​​the vehicle, C D is the aerodynamic drag coefficient of the vehicle, η M , η D They are the efficiency of the vehicle motor and the transmission system, ΔO ij is the driving loss of road ij, is the distance of the kth journey, C B , η B are battery capacity, battery efficiency, and O I , O F are the starting SOC and the remaining SOC, n N is the number of network nodes, is the intermediate parameter; For the starting time, the starting time of buses is set to 07:30; the starting time of private cars and taxis is determined according to the following probability density function: In the above formula, t s is the starting travel time, α is the scale parameter, C and K are shape parameters; For parking time, the parking time after a single trip chain for buses is set at 30 minutes; the parking time for private cars and taxis varies according to the location; For the starting SOC, the starting SOC of the bus is set to 1; the starting SOC of private cars and taxis is determined according to the following probability density function: In the above formula, O I is the starting SOC, σ soc , μ soc are the mean and standard deviation of the Gaussian distribution.

4. A method for predicting electric vehicle charging demand by distinguishing vehicle types according to claim 1 or 2, characterized in that: In S1, the transfer matrix is: G={N, R, T, W, D, Q, S, V} N={o i |i=1,2,…,n N } R={(o i ,o j )|i∈N,j∈N,i≠j} T={t|t=1,2,…,24} D={d ij |(o i ,o j )∈R} In the above formula, G is the set of traffic network models, N is the set of intersection nodes in the network, R is the set of roads in the network, T is the set of time periods, W is the set of road weights in the network, D is the set of actual distances between nodes, Q is the set of road traffic volume in the network, S is the set of road saturation, V is the set of road speeds, and n is the set of road traffic volume. N is the number of network nodes, o i , o j are the i-th and j-th nodes respectively, t is the time period number, is the road weight of road ij in period t, d ij is the length of road ij, are the traffic volume, saturation and average speed of road ij in period t respectively.

5. The method for predicting charging demand of electric vehicles by differentiating vehicle types according to claim 1 or 2, characterized in that: The S2 uses the Dijkstra algorithm to determine the optimal driving path of the electric vehicle.

6. A system for predicting charging demand of electric vehicles by distinguishing vehicle types, characterized in that: The system includes a travel start point and end point determination module, an optimal driving path calculation module, a remaining SOC calculation module, a first judgment module, a second judgment module, and a third judgment module; The travel start point and end point determination module is used to determine the travel start point and end point of the electric vehicle through a transfer matrix; The optimal driving path calculation module is used to determine the optimal driving path of the electric vehicle based on the travel starting point, the end point and the road weight; The remaining SOC calculation module is used to consider different types of electric vehicles and calculate the remaining SOC of the electric vehicle at each node based on the optimal driving path using an established energy consumption model; The first judgment module is used to judge whether the electric vehicle has a charging demand according to the remaining SOC at each node. If so, the second judgment module is started after the charging amount of the node is updated; if not, the second judgment module is directly started; The second judgment module is used to judge whether the trip destination has been reached. If not, the remaining SOC, the start time and location of the next trip are updated and sent to the travel start and end point determination module for the next cycle; if it has been reached, the third judgment module is started; The third judgment module is used to judge whether all electric vehicles have completed their journeys. If so, the spatiotemporal distribution results of the electric vehicle charging requirements are generated. If not, the travel start and end point determination module is started to calculate the journey of the next electric vehicle until the charging requirements of all electric vehicles are predicted.

7. The electric vehicle charging demand prediction system for distinguishing vehicle types according to claim 1 is characterized in that: The road weight is calculated according to the following formula: In the above formula, is the road weight of road ij in time period t, ρ1 and ρ2 are the weight factors of road impedance and road section distance respectively, d ij is the length of road ij, R ij (t) is the road impedance of road ij in period t, t0 is the travel time under zero flow conditions, l and ε are impedance influencing factors, is the saturation of road ij in period t, are the traffic volume and road capacity of road ij in period t respectively.

8. The electric vehicle charging demand prediction system according to claim 6 or 7, characterized in that: The energy consumption model includes: In the above formula, is the power required by the tires of an electric vehicle when it travels on road ij, c r , c1, c2 are rolling resistance coefficients, m and g are vehicle mass and gravity acceleration respectively. is the average speed of road ij in period t, ρ a is the air density, A f is the front area of ​​the vehicle, C D is the aerodynamic drag coefficient of the vehicle, η M , η D They are the efficiency of the vehicle motor and the transmission system, ΔO ij is the driving loss of road ij, is the distance of the kth journey, C B , η B are battery capacity, battery efficiency, and O I , O F are the starting SOC and the remaining SOC, n N is the number of network nodes, is the intermediate parameter; For the starting time, the starting time of buses is set to 07:30; the starting time of private cars and taxis is determined according to the following probability density function: In the above formula, t s is the starting travel time, α is the scale parameter, C and K are shape parameters; For parking time, the parking time after a single trip chain for buses is set at 30 minutes; the parking time for private cars and taxis varies according to the location; For the starting SOC, the starting SOC of the bus is set to 1; the starting SOC of private cars and taxis is determined according to the following probability density function: In the above formula, O I is the starting SOC, σ soc , μ soc are the mean and standard deviation of the Gaussian distribution.

9. The electric vehicle charging demand prediction system according to claim 6 or 7, characterized in that: The transfer matrix is: G={N, R, T, W, D, Q, S, V} N={o i |i=1,2,…,n N } R={(o i ,o j )|i∈N,j∈N,i≠j} T={t|t=1,2,…,24} D={d ij |(o i ,o j )∈R} In the above formula, G is the set of traffic network models, N is the set of intersection nodes in the network, R is the set of roads in the network, T is the set of time periods, W is the set of road weights in the network, D is the set of actual distances between nodes, Q is the set of road traffic volume in the network, S is the set of road saturation, V is the set of road speeds, and n is the set of road traffic volume. N is the number of network nodes, o i , o j are the i-th and j-th nodes respectively, t is the time period number, is the road weight of road ij in period t, d ij is the length of road ij, are the traffic volume, saturation and average speed of road ij in period t respectively.

10. The electric vehicle charging demand prediction system for distinguishing vehicle types according to claim 6 or 7, characterized in that: The optimal driving path calculation module uses the Dijkstra algorithm to determine the optimal driving path of the electric vehicle.

Images