Electric vehicle virtual energy storage space-time distribution prediction method and system
Through the Logit-based random allocation model and real-time traffic flow update, the problem of insufficient user behavior and traffic flow dynamic characteristics in the virtual energy storage prediction of electric vehicles is solved, and higher-precision prediction and grid auxiliary services are achieved.
Patent Information
- Application Number
- CN202510595690.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-09
- Publication Date
- 2025-07-22
AI Technical Summary
The existing virtual energy storage status prediction methods for electric vehicles are difficult to accurately reflect the user's actual travel behavior and time-varying road traffic flow distribution, resulting in insufficient prediction accuracy.
The random allocation model based on Logit is used to combine road traffic flow to determine the travel path of electric vehicles and update the traffic flow in real time to calculate the state of charge and charging needs, taking into account the user's limited rationality and traffic flow changes.
It improves the accuracy of the spatial and temporal distribution prediction of virtual energy storage for electric vehicles, enhances the operation flexibility of the power grid and the ability of electric vehicles to participate in grid auxiliary services.
Smart Images
Figure CN120355035A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of electric vehicles, and in particular to a method and system for predicting the spatiotemporal distribution of virtual energy storage in electric vehicles. Background Art
[0002] In order to achieve carbon neutrality as soon as possible and promote the access of new energy, the transformation of transportation electrification is accelerating, and the number of electric vehicles is increasing year by year, and it is expected to reach 80 million in 2030. Electric vehicles not only have transportation attributes, but also energy attributes. The aggregated electric vehicle cluster can participate in the auxiliary services of the power grid as a virtual energy storage, providing flexibility support for the power grid. Therefore, accurately predicting the spatiotemporal distribution of the virtual energy storage state of electric vehicles is helpful for the optimized operation of the power grid.
[0003] Since the virtual energy storage state of electric vehicles is closely related to their charging behavior, the method of electric vehicle charging load prediction can be used to predict the virtual energy storage state. From the current research on charging load prediction, some of them plan the user's travel path with the shortest travel path as the goal, and some plan the travel path with the shortest driving time as the goal, but it is difficult to reflect the actual travel behavior of users. Although some studies have proposed a charging load prediction method based on random user equilibrium and travel chain, its analysis step size is difficult to reflect the time-varying road traffic flow distribution. If the impact of user limited rationality and time-varying traffic flow on travel decisions can be considered, it will help improve the accuracy of electric vehicle virtual energy storage prediction. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a method and system for predicting the spatiotemporal distribution of virtual energy storage of electric vehicles, which can improve the accuracy of prediction.
[0005] In order to solve the above technical problems, a technical solution adopted by the present invention is: A method for predicting the spatiotemporal distribution of virtual energy storage of electric vehicles, comprising the steps of: Determine the current electric private car from all electric private cars at the current road network node; If the current electric private car is in a driving state, obtaining a set of feasible paths from a current road network node of the current electric private car to a destination road network node; Determine a travel path from the set of feasible paths based on road traffic flow using a Logit-based random assignment model, and update the road traffic flow; If the current electric private car arrives at the destination road network node based on the travel path, then calculating the charge state of the current electric private car when it arrives at the destination; Determine whether the state of charge is less than or equal to the charging threshold. If so, calculate the charging demand and virtual energy storage state of the current electric private car according to the state of charge, and after determining the next electric private car from all the electric private cars as the current electric private car, return to execute the step of "if the current electric private car is in a driving state".
[0006] To solve the above technical problems, another technical solution adopted by the present invention is: An electric vehicle virtual energy storage spatio-temporal distribution prediction system, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented: Determine the current electric private car from all the electric private cars at the current road network node; If the current electric private car is in a driving state, obtain a set of feasible paths from the current road network node of the current electric private car to the destination road network node; Use a Logit-based stochastic assignment model to determine the travel path from the set of feasible paths based on the road traffic flow, and update the road traffic flow; If the current electric private car arrives at the destination road network node based on the travel path, calculate the state of charge of the current electric private car when it arrives at the destination; Determine whether the state of charge is less than or equal to the charging threshold. If so, calculate the charging demand and virtual energy storage state of the current electric private car according to the state of charge, and after determining the next electric private car from all the electric private cars as the current electric private car, return to execute the step of "if the current electric private car is in a driving state".
[0007] The beneficial effects of the present invention are as follows: If the current electric private car is in a driving state, obtain a set of feasible paths from the current road network node of the current electric private car to the destination road network node, use a Logit-based stochastic assignment model to determine the travel path from the set of feasible paths based on the road traffic flow, and update the road traffic flow. If the current electric private car arrives at the destination road network node based on the travel path, calculate the state of charge of the current electric private car when it arrives at the destination. If the state of charge is less than or equal to the charging threshold, calculate the charging demand and virtual energy storage state of the current electric private car according to the state of charge. In this way, during the process of predicting the spatio-temporal distribution of electric vehicle virtual energy storage, the Logit-based stochastic assignment model is used to consider the bounded rationality of users, and the impact of traffic flow changes on travel decisions is considered by updating the road traffic flow, overcoming the deficiencies of users' complete rationality and ignoring the impact of the dynamic characteristics of the traffic network on users' travel, thereby improving the prediction accuracy and contributing to the participation of electric vehicle virtual energy storage in the ancillary services of the power grid and improving the operation flexibility of the power grid. Description of the Drawings
[0008] Figure 1 It is a flowchart of the steps of a method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to an embodiment of the present invention; Figure 2 It is a schematic structural diagram of a system for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to an embodiment of the present invention; Figure 3 It is a prediction framework diagram in the method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to an embodiment of the present invention; Figure 4 It is an actual traffic network diagram in the method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to an embodiment of the present invention; Figure 5 It is a calculation flowchart in the method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to an embodiment of the present invention; Figure 6 It is a schematic diagram of the spatio-temporal distribution of virtual energy storage of electric vehicles in the method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to an embodiment of the present invention; Figure 7 It is a schematic diagram of the spatio-temporal distribution of road traffic flow in the method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to an embodiment of the present invention; Figure 8 It is a schematic diagram of the spatio-temporal distribution of the charging load of electric vehicles in the method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to an embodiment of the present invention. Detailed Embodiments
[0009] To describe in detail the technical content, achieved objectives and effects of the present invention, the following is described in conjunction with the embodiments and with reference to the drawings.
[0010] Please refer to Figure 1 , a method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles, comprising the steps of: Determine the current electric private car from all the electric private cars at the current road network node; If the current electric private car is in a driving state, obtain a set of feasible paths from the current road network node of the current electric private car to the destination road network node; Use the Logit-based stochastic assignment model to determine the travel path from the set of feasible paths based on the road traffic flow, and update the road traffic flow; If the current electric private car arrives at the destination road network node based on the travel path, calculate the state of charge of the current electric private car when it arrives at the destination; Determine whether the state of charge is less than or equal to the charging threshold. If so, calculate the charging demand and the virtual energy storage state of the current electric private car according to the state of charge. After determining the next electric private car from all the electric private cars as the current electric private car, return to execute the step of "if the current electric private car is in a driving state".
[0011] As can be seen from the above description, the beneficial effects of the present invention are as follows: If the current electric private car is in a driving state, obtain the set of feasible paths from the current road network node of the current electric private car to the destination road network node, use the Logit-based stochastic assignment model to determine the travel path from the set of feasible paths based on the road traffic flow, and update the road traffic flow. If the current electric private car arrives at the destination road network node based on the travel path, calculate the state of charge of the current electric private car when it arrives at the destination. If the state of charge is less than or equal to the charging threshold, calculate the charging demand and the virtual energy storage state of the current electric private car according to the state of charge. In this way, in the process of predicting the spatio-temporal distribution of the virtual energy storage of electric vehicles, the Logit-based stochastic assignment model is used to consider the bounded rationality of users, and the impact of traffic flow changes on travel decisions is considered by updating the road traffic flow, overcoming the deficiencies of users' perfect rationality and ignoring the impact of the dynamic characteristics of the traffic network on users' travel, thereby improving the prediction accuracy and helping the virtual energy storage of electric vehicles to participate in the ancillary services of the power grid and improving the operation flexibility of the power grid.
[0012] Furthermore, it also includes: If the current electric private car is in a parked state, calculate the state of charge of the current electric private car; Determine whether the state of charge is less than or equal to the charging threshold. If not, after determining the next electric private car from all the electric private cars as the current electric private car, return to execute the step of "if the current electric private car is in a driving state". If so, calculate the charging demand and the virtual energy storage state of the current electric private car according to the state of charge. After determining the next electric private car from all the electric private cars as the current electric private car, return to execute the step of "if the current electric private car is in a driving state".
[0013] As can be seen from the above description, for vehicles in a parked state with a state of charge lower than the threshold, while calculating their charging demand, their potential as virtual energy storage resources is also evaluated, which helps to better integrate electric vehicles into the power grid energy storage system and realize the vehicle-to-grid (V2G) function.
[0014] Furthermore, it also includes: If the current electric private car has not reached the destination road network node based on the travel route, update the current road network node of the current electric private car, and after taking the next electric private car determined from all the electric private cars as the current electric private car, return to execute the step of if the current electric private car is in a driving state.
[0015] As can be seen from the above description, when the electric private car has not reached the destination, by continuously updating the road network node of the current electric private car, the system can accurately grasp the precise position of the vehicle in real time, reasonably arrange the vehicle to participate in the auxiliary services of the power grid, and provide flexibility support for the power grid.
[0016] Furthermore, it further includes: If the state of charge is greater than the charging threshold, after taking the next electric private car determined from all the electric private cars as the current electric private car, return to execute the step of if the current electric private car is in a driving state.
[0017] As can be seen from the above description, when the state of charge is greater than the charging threshold, it indicates that the electric private car does not need to be charged, avoiding waste of charging resources, and the next electric private car can be processed.
[0018] Furthermore, the use of the Logit-based stochastic assignment model to determine the travel route from the set of feasible routes based on road traffic flow includes: ; ; In the formula, represents the probability of selecting the route t in the set of feasible routes between the departure place r and the destination s during the time period k , represents the user's degree of mastery of the traffic network condition, represents the travel cost of the route t during the time period k , K rs represents the set of feasible routes between the departure place r and the destination s , x ij,t represents the traffic flow of the road ( i , j ) during the time period t , t ( x ij,t ) represents the travel time of the road ( i , j ) during the time period t . Indicates the path k The correlation coefficient between the path and the road( i , j ).
[0019] As can be seen from the above description, when using the Logit-based stochastic assignment model to determine the travel path, the bounded rationality of users is considered. The electric private car does not always aim to minimize the travel time when choosing a travel path, which can effectively improve the prediction accuracy of the spatio-temporal distribution of the virtual energy storage of electric vehicles.
[0020] Furthermore, the updating of the road traffic flow includes: ; In the formula, t ( x ij,t ) represents the travel time of the road( i , j ) during the time period t , x ij,t represents the traffic flow of the road( i , j ) during the time period t , x ij,t-1 represents the traffic flow of the road( i , j ) during the time period t -1, represents the free travel time of the road( i , j ), represents the first road parameter, cap ij represents the capacity of the road( i , j ), represents the second road parameter.
[0021] As can be seen from the above description, since the travel time is the main factor considered by users when choosing a path, real-time updating of the road traffic flow effectively improves the accuracy of the final prediction.
[0022] Furthermore, the calculation of the state of charge of the current electric private car when it arrives at the destination includes: ; In the formula, SOC arr represents the state of charge of the current electric private car when it arrives at the destination, SOC dep represents the state of charge of the current electric private car when it departs, l rsIndicates the driving distance, EC indicates the driving energy consumption, and BC indicates the battery capacity.
[0023] As can be seen from the above description, by calculating the state of charge of the current electric private car when it reaches the destination and calculating the charging demand of the electric vehicle and the spatio-temporal distribution of virtual energy storage based on the travel chain and charging characteristics, it is more accurate and reliable.
[0024] Furthermore, the calculating the charging demand of the current electric private car and the virtual energy storage state according to the state of charge includes: ; ; In the formula, P n,t Indicates the charging demand of the road network node n At time t The charging demand at V c Indicates the time t The number of electric private cars in the charging state at the road network node n ; C power Indicates the charging power E n,t Indicates the road network node n At time t The virtual energy storage state at V p Indicates the time t The number of electric private cars in the parked state at the road network node n ; Indicates the time t The road network node n In the v p The state of charge of the
[0025] As can be seen from the above description, calculating the charging demand of the electric private car and the virtual energy storage state takes into account the state of charge, ensuring the reliability and accuracy of the prediction.
[0026] Furthermore, after determining the next electric private car from all the electric private cars as the current electric private car, the returning to execute the step of if the current electric private car is in the driving state includes: Determining the next electric private car that has not been used as the current electric private car from all the electric private cars; Taking the next electric private car as the current electric private car and returning to execute the step of if the current electric private car is in the driving state; If all the electric private cars have been regarded as the current electric private car, it is judged whether there is a next road network node. If so, after taking the next road network node as the current road network node, return to execute the step of determining the current electric private car from all the electric private cars at the current road network node. If not, obtain the travel trajectories of fuel vehicles and electric taxis, and update the road traffic flow according to the travel trajectories of the fuel vehicles and electric taxis. Increment the current time by one, and judge whether the incremented current time reaches the preset time. If so, output the virtual energy storage state and the charging demand at each time and the road traffic flow in each time period. If not, update the road travel time according to the road traffic flow, and return to execute the step of determining the current electric private car from all the electric private cars at the current road network node.
[0027] As can be seen from the above description, predicting the spatio-temporal distribution of virtual energy storage for each electric private car in each road network node helps the virtual energy storage of electric vehicles to participate in grid auxiliary services and traffic-power coupling.
[0028] Please refer to Figure 2 , another embodiment of the present invention provides an electric vehicle virtual energy storage spatio-temporal distribution prediction system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, each step in the above-mentioned electric vehicle virtual energy storage spatio-temporal distribution prediction method is implemented.
[0029] The above-mentioned electric vehicle virtual energy storage spatio-temporal distribution prediction method and system of the present invention can be applied to electric vehicles, which will be described below through specific embodiments: Please refer to Figure 1 , Figures 3 - 8 , the first embodiment of the present invention is: An electric vehicle virtual energy storage spatio-temporal distribution prediction method, including the steps: S1. Determine the current electric private car from all the electric private cars at the current road network node. In an optional implementation manner, initialize the current time t = 1, and the current electric private car i = 1.
[0030] Among them, electric vehicles can be divided into electric private cars and electric taxis. Since electric vehicles only participate in grid operation as virtual energy storage when they are in a parked state, and require a long idle time and quantity, the present invention only regards electric private cars as the research object of virtual energy storage, and fuel vehicles and electric taxis only affect the distribution of road traffic flow.
[0031] In an optional implementation manner, before S1, it further includes: Determine the current road network node from all road network nodes. In an alternative embodiment, initialize the current road network node n = 1.
[0032] In an alternative embodiment, as Figure 3 shown, before S1, it further includes: Obtain the road length, design speed, capacity, traffic network topology information, including the number of road network nodes N , the number of edges, the connection relationship between road network nodes, the battery capacity, energy consumption rate, charging power, charging threshold of electric vehicles, the travel chain, departure location, initial state of charge, first departure time, destination parking duration (including office area parking duration and commercial area parking duration) of each electric private car, and the initial number of electric private cars at each road network node I n , the number of electric taxis, the actual OD demand of the road network (traffic travel demand between the origin and the destination). Among them, the actual traffic network is as Figure 4 shown.
[0033] For electric private car users, the travel chain method is used to describe their travel trajectories. The travel destinations of electric taxi users are randomly generated, and electric taxis are more sensitive to travel time. It is considered that electric taxis choose travel paths with the shortest travel time as the goal. The travel chains of electric private cars on weekdays are relatively simple. According to the number of destinations, they are mainly divided into two types. One travel chain takes the office area as the destination and returns home after work. The other travel chain considers the shopping or entertainment behavior of users after work and adds the commercial area as one of the destinations.
[0034] Among them, the first departure time, the office area parking duration, and the commercial area parking duration of the electric private car are calculated and obtained through the following formulas: ; ; ; In the formula, f ( t dep ) represents the probability density distribution function of the first departure time, t dep represents the first departure time, represents the first weight, represents the first standard deviation, represents the first mean, f ( ) represents the probability density distribution function of the office area parking duration, Indicates the parking duration in the office area, Indicates the second weight, Indicates the second standard deviation, Indicates the second mean value, f ( ) represents the probability density distribution function of the parking duration in the commercial area, Indicates the parking duration in the commercial area, Indicates the third weight, Indicates the third standard deviation, Indicates the third mean value.
[0035] S2. If the current electric private car i is in the driving state, then execute S4 - S8, as Figure 5 shown.
[0036] S3. If the current electric private car i is in the parked state, then execute S9 - S10, as Figure 5 shown.
[0037] S4. Obtain the set of feasible paths from the current road network node of the current electric private car i to the destination road network node.
[0038] In an alternative embodiment, use the k - shortest path algorithm based on Dijkstra to obtain the set of feasible paths from the current road network node of the current electric private car i to the destination road network node.
[0039] S5. Use the Logit - based stochastic assignment model to determine the travel path from the set of feasible paths based on the road traffic flow and update the road traffic flow.
[0040] Among them, the use of the Logit - based stochastic assignment model to determine the travel path from the set of feasible paths based on the road traffic flow includes: ; ; In the formula, represents the time period t within which the probability of selecting path r from the set of feasible paths between the departure place s and the destination k is, represents the user's understanding of the traffic network conditions, represents the time period t within which the travel cost of path k is, K rs represents the departure placer The set of feasible paths between the origin and the destination s is denoted as x ij,t where i , j represents the traffic flow on the road t during the time period t ( x ij,t ) represents the travel time of the road i , j during the time period t , denotes the correlation coefficient between the path k and the road i , j ).
[0041] The update of the traffic flow on the road includes: ; In the formula, x ij,t-1 represents the traffic flow on the road i , j during the time period t -1, represents the free travel time of the road i , j , represents the first road parameter, cap ij represents the capacity of the road i , j , represents the second road parameter.
[0042] In the above formula, the travel time of the road is calculated using BPR (Bayesian Posterior Optimization-based Personalized Ranking Algorithm) and regarded as the travel cost.
[0043] Among them, for electric private cars, when the origin is home, the destination arrival time is specifically: ; In the formula, t arr represents the destination arrival time, t rs represents the travel time from the origin to the destination.
[0044] When the origin is the office area or the business area, the destination arrival time is specifically: ; ; In the formula, tleave Indicates the departure time from the previous destination, t arr ′ indicates the arrival time at the previous destination.
[0045] S6. If the current electric private car i arrives at the destination road network node based on the travel route, then calculate the state of charge of the current electric private car i when it arrives at the destination, specifically: ; In the formula, SOC arr represents the state of charge of the current electric private car when it arrives at the destination, SOC dep represents the state of charge of the current electric private car when it departs, l rs represents the driving distance, EC represents the driving energy consumption, and BC represents the battery capacity.
[0046] Among them, when the departure place of the electric private car is home, according to whether the initial state of charge SOC ini reaches the charging threshold, SOC dep is divided into two cases, specifically: ; In the formula, t c represents the charging duration, SOC thr represents the charging threshold.
[0047] When the departure place of the electric private car is other attribute areas, SOC dep is divided into two cases, specifically: ; In the formula, SOC arr ′ represents the state of charge when arriving at the previous destination.
[0048] S7. Judge whether the state of charge is less than or equal to the charging threshold. If so, calculate the charging demand and virtual energy storage state of the current electric private car according to the state of charge, and after determining the next electric private car from all the electric private cars as the current electric private car, return to execute S2. If not, after determining the next electric private car from all the electric private cars as the current electric private car, return to execute S2. i
[0049] Among them, calculating the charging demand and virtual energy storage state of the current electric private car according to the state of charge includes: i ; ; ; In the formula, P n,t represents the charging demand of the road network node n at time t , V c represents the number of electric private cars in the charging state in the road network node t at time n , C power represents the charging power, E n,t represents the virtual energy storage state of the road network node n at time t , V p represents the number of electric private cars in the parked state in the road network node t at time n , represents the state of charge of the t th n electric private car in the road network node v p at time
[0050] S8. If the current electric private car i does not reach the destination road network node based on the travel path, update the current road network node of the current electric private car i , and after determining the next electric private car from all the electric private cars as the current electric private car, return to execute S2.
[0051] Among them, updating the current road network node of the current electric private car i specifically means: updating the current road network node of the current electric private car i at time t .
[0052] S9. Calculate the state of charge of the current electric private car.
[0053] S10. Judge whether the state of charge is less than or equal to the charging threshold. If not, determine the next electric private car from all the electric private cars as the current electric private car and then return to execute S2. If so, calculate the current electric private car according to the state of charge ithe charging demand and the virtual energy storage state, and after determining the next electric private car from all the electric private cars as the current electric private car, return to execute S2.
[0054] Among them, after determining the next electric private car from all the electric private cars as the current electric private car and returning to execute S2, it includes: Determine the next electric private car that has not been used as the current electric private car from all the electric private cars; Take the next electric private car as the current electric private car and return to execute S2; If all the electric private cars have been used as the current electric private car, determine whether there is a next road network node. If there is, take the next road network node as the current road network node and return to execute S1. If not, obtain the travel trajectories of fuel vehicles and electric taxis, and update the road traffic flow according to the travel trajectories of the fuel vehicles and electric taxis; Increment the current time by one, and determine whether the incremented current time reaches the preset time. If so, output the virtual energy storage state and the charging demand at each time and the road traffic flow in each time period. If not, update the road travel time according to the road traffic flow and return to execute S1.
[0055] Specifically, let i = i +1, and determine whether i is less than or equal to I n . If so, return to execute S2. If not, let n = n +1, and determine whether n is less than N . If it is less than N , let i =1, and return to execute S1. If it is not less than N , obtain the travel trajectories of fuel vehicles and electric taxis, and update the road traffic flow according to the travel trajectories of the fuel vehicles and electric taxis; let t = t +1, and determine whether t is equal to the preset time. If it is not equal, update the road travel time according to the road traffic flow and return to execute S1. Let n =1, i =1. If it is equal, end the calculation process and output the virtual energy storage state and the charging demand at each time and the road traffic flow in each time period. In an optional implementation manner, the preset time is 288.
[0056] Such as Figure 6 ,Figure 7 and Figure 8 as shown Figure 6 shows the spatio-temporal distribution of virtual energy storage of electric vehicles Figure 7 shows the spatio-temporal distribution of road traffic flow Figure 8 shows the spatio-temporal distribution of charging load of electric vehicles
[0057] Please refer to Figure 2 , Embodiment 2 of the present invention is as follows: A spatio-temporal distribution prediction system for virtual energy storage of electric vehicles, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, each step in the spatio-temporal distribution prediction method of virtual energy storage of electric vehicles in Embodiment 1 is implemented.
[0058] In summary, the present invention provides a spatio-temporal distribution prediction method and system for virtual energy storage of electric vehicles. If the current electric private vehicle is in a driving state, a set of feasible paths from the current road network node of the current electric private vehicle to the destination road network node is obtained, and a travel path is determined from the set of feasible paths based on the road traffic flow using a Logit-based stochastic assignment model, and the road traffic flow is updated. If the current electric private vehicle arrives at the destination road network node based on the travel path, the state of charge when the current electric private vehicle arrives at the destination is calculated. If the state of charge is less than or equal to the charging threshold, the charging demand and the virtual energy storage state of the current electric private vehicle are calculated according to the state of charge. In this way, during the process of predicting the spatio-temporal distribution of virtual energy storage of electric vehicles, the Logit-based stochastic assignment model is used to consider the bounded rationality of users, and the impact of traffic flow changes on travel decisions is considered by updating the road traffic flow, overcoming the deficiencies of users' complete rationality and ignoring the impact of the dynamic characteristics of the traffic network on users' travel, thereby improving the prediction accuracy, contributing to the participation of virtual energy storage of electric vehicles in the ancillary services of the power grid, and improving the operation flexibility of the power grid; in addition, for vehicles in a parked state and with a state of charge lower than the threshold, while calculating their charging demand, their potential as virtual energy storage resources is also evaluated, which helps to better integrate electric vehicles into the power grid energy storage system and realize the vehicle-to-grid function.
[0059] The above are only the embodiments of the present invention, and do not limit the patent scope of the present invention. All equivalent transformations made using the content of the specification and drawings of the present invention, or directly or indirectly applied in related technical fields, are equally included in the patent protection scope of the present invention.
Claims
1. A prediction method for the spatio-temporal distribution of virtual energy storage in electric vehicles, characterized in that, Including the steps: Determine the current electric private car from all the electric private cars at the current road network node; If the current electric private car is in a driving state, obtain the set of feasible paths from the current road network node of the current electric private car to the destination road network node; Use the Logit-based stochastic assignment model to determine the travel path from the set of feasible paths based on the road traffic flow, and update the road traffic flow; If the current electric private car arrives at the destination road network node based on the travel path, calculate the state of charge of the current electric private car when it arrives at the destination; Judge whether the state of charge is less than or equal to the charging threshold. If so, calculate the charging demand and virtual energy storage state of the current electric private car according to the state of charge, and after taking the next electric private car determined from all the electric private cars as the current electric private car, return to execute the step of if the current electric private car is in a driving state; 2. The virtual energy storage spatio-temporal distribution prediction method for an electric vehicle according to claim 1, characterized in that It also includes: If the current electric private car is in a parked state, calculate the state of charge of the current electric private car; Judge whether the state of charge is less than or equal to the charging threshold. If not, after taking the next electric private car determined from all the electric private cars as the current electric private car, return to execute the step of if the current electric private car is in a driving state. If so, calculate the charging demand and virtual energy storage state of the current electric private car according to the state of charge, and after taking the next electric private car determined from all the electric private cars as the current electric private car, return to execute the step of if the current electric private car is in a driving state; 3. A method for predicting the spatio-temporal distribution of virtual energy storage in an electric vehicle, according to claim 1, characterized in that It also includes: If the current electric private car does not arrive at the destination road network node based on the travel path, update the current road network node of the current electric private car, and after taking the next electric private car determined from all the electric private cars as the current electric private car, return to execute the step of if the current electric private car is in a driving state; 4. A method for predicting the spatio-temporal distribution of virtual energy storage in an electric vehicle according to claim 1, characterized in that, It also includes: If the state of charge is greater than the charging threshold, after taking the next electric private car determined from all the electric private cars as the current electric private car, return to execute the step of if the current electric private car is in a driving state; 5. A method for predicting the spatio-temporal distribution of virtual energy storage in an electric vehicle, according to claim 1, wherein The using the Logit-based stochastic assignment model to determine the travel path from the set of feasible paths based on the road traffic flow includes: ； ； wherein, represents a time period t within the origin r and the destination s selecting a path from the set of feasible paths k with probability, represents the user's degree of understanding of the traffic network conditions, represents a time period t the cost of passing through the path k during this period, K rs represents the origin r and the destination s the set of feasible paths between, x ij,t represents the road( i , j ) during the time period t traffic flow, t ( x ij,t ) represents the road( i , j ) during the time period t travel time, represents the path k and the road( i , j ) the correlation coefficient between.
6. The virtual energy storage spatio-temporal distribution prediction method for an electric vehicle according to claim 1, characterized in that The updating the road traffic flow includes: ； Wherein, t ( x ij,t ) represents the passing time of road ( i , j ) during time period t . x ij,t represents the traffic flow of road ( i , j ) during time period t . x ij,t-1 represents the traffic flow of road ( i , j ) during time period t - 1. represents the free passing time of road ( i , j ). represents the first road parameter. cap ij represents the capacity of road ( i , j ). represents the second road parameter.
7. A method for predicting the spatio-temporal distribution of virtual energy storage in an electric vehicle, according to claim 1, characterized in that The calculating the state of charge of the current electric private car when it arrives at the destination includes: ； Wherein, SOC arr represents the state of charge when the current electric private car arrives at the destination, SOC dep represents the state of charge when the current electric private car departs, l rs represents the driving distance, EC represents the driving energy consumption, and BC represents the battery capacity.
8. A method for predicting the spatio-temporal distribution of virtual energy storage of an electric vehicle according to claim 1, characterized in that The calculating the charging demand and virtual energy storage state of the current electric private car according to the state of charge includes: ； ； In the formula, P n,t represents the road network node n at time t 's charging demand, V c represents the time t the number of electric private cars in the charging state in the road network node n at C power represents the charging power, E n,t represents the road network node n at time t 's virtual energy storage state, V p represents the time t the number of electric private cars in the parked state in the road network node n at represents the time t the road network node n in the v p th electric private car's state of charge.
9. A method for predicting the spatio-temporal distribution of virtual energy storage in an electric vehicle, according to any one of claims 1 to 4, characterized in that The after taking the next electric private car determined from all the electric private cars as the current electric private car and returning to execute the step of if the current electric private car is in a driving state includes: Determine the next electric private car that has not been used as the current electric private car from all the electric private cars; Take the next electric private car as the current electric private car, and return to execute the step of if the current electric private car is in a driving state; If all electric private cars have been taken as the current electric private car, determine whether there is a next road network node. If there is, take the next road network node as the current road network node, and then return to execute the step of determining the current electric private car from all electric private cars at the current road network node. If not, obtain the travel trajectories of fuel vehicles and electric taxis, and update the road traffic flow according to the travel trajectories of the fuel vehicles and electric taxis; Increment the current time by one, and determine whether the incremented current time reaches the preset time. If so, output the virtual energy storage state and the charging demand at each time and the road traffic flow in each time period. If not, update the road travel time according to the road traffic flow, and return to execute the step of determining the current electric private car from all electric private cars at the current road network node; 10. A virtual energy storage spatio-temporal distribution prediction system for an electric vehicle, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements each step in a method for predicting the spatio-temporal distribution of virtual energy storage of electric vehicles according to any one of claims 1 to 9.