Electric vehicle dynamic congestion charging scheduling method based on multi-level target planning

Abstract

The invention discloses an electric vehicle dynamic congestion charging scheduling method based on multi-level target planning. The method comprises the following steps: establishing an electric vehicle charging demand response model; establishing an urban traffic dynamic congestion model of the traffic network and the charging and discharging dynamic congestion degree; based on the charging and discharging behavior characteristics of the user, establishing an electric vehicle charging and discharging model for quantifying the energy interaction process between the electric vehicle and the charging station and the energy consumption in the process of going to the charging station; establishing a multi-level target planning model, and constructing a decision-making hierarchical structure; all the charging stations are listed as alternative schemes by taking the selection of the optimal charging station as the target and taking the maximization of the energy quantity requested by the electric vehicle, the minimization of the total response time and the minimization of the charging cost as the criteria, the priority weights of the criteria and the alternative schemes are compared and determined, and the weighted deviation sum of each target is minimized. The method realizes user personalized decision, meets diversified requirements, improves the stability of the power distribution network, and improves the carrying capacity of the power distribution network to the electric vehicle.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle charging scheduling technology, and specifically to a dynamic congestion charging scheduling method for electric vehicles based on multi-level target planning. Background Technology

[0002] In recent years, my country's new energy vehicle market has developed rapidly. Electric vehicles, as flexible loads at the user end of the distribution network, have begun to participate in flexible power consumption regulation on the load side, thereby changing the load conditions of the power grid. However, their charging and discharging processes are highly uncertain and vary regionally. The disorderly charging of a large number of electric vehicles can cause drastic fluctuations in the charging load of the power grid, leading to an exacerbation of the peak-valley difference in the distribution network and difficulties in dispatching. Distribution network dispatching for electric vehicles is increasingly becoming the core link of the "source-grid-load-storage" synergy in the new power system, transforming the dispersed electric vehicles into programmable grid assets and achieving a dynamic balance between safety, economy, and low carbon goals.

[0003] Electric vehicle (EV) charging loads are transferable, and adopting reasonable charging strategies can significantly improve the flexibility and economy of the distribution network. EV charging stations, acting as EV aggregators, can leverage the flexibility of their charging load regulation to guide users to distribute charging demand through peak-valley pricing or incentive mechanisms, thereby reducing the maximum load on the distribution network, decreasing equipment investment costs, and generating profits in the ancillary services market. Furthermore, as microgrid nodes, charging stations participate in local energy balancing, supporting critical load power supply in off-grid mode and enhancing the resilience of the distribution network. However, existing research only considers single factors such as price or charging cost, failing to account for the combined impact of multiple factors, including charging time. Summary of the Invention

[0004] The purpose of this invention is to address the problems that have arisen in the process of technological development. Specifically, it proposes a dynamic congestion-based charging scheduling method for electric vehicles (EVs) based on multi-level objective programming, which addresses the issues of current EV charging scheduling strategies being too simplistic and lacking dynamic response capabilities. This method dynamically adjusts objective weights and allocation strategies through multi-objective dynamic optimization and weight adaptation, thereby adapting to real-time changes in traffic, power grid, and user demand. It enables personalized user decision-making to meet diverse needs, improves the stability of the power distribution network, and increases the network's capacity to support EVs.

[0005] The technical solution adopted in this invention is as follows:

[0006] A dynamic congestion-based charging scheduling method for electric vehicles based on multi-level objective programming includes the following steps:

[0007] Step 1: Establish an electric vehicle charging demand response model based on the characteristics of electric vehicle user behavior, driving time, and electric vehicle charging waiting time costs;

[0008] Step 2: Based on the traffic flow dynamics and user charging / discharging density in different urban areas, establish an urban traffic dynamic congestion model that integrates the traffic network with the dynamic congestion of charging / discharging, thereby achieving accurate quantification of road travel time;

[0009] Step 3: Based on the user's charging and discharging behavior characteristics, establish an electric vehicle charging and discharging model to quantify the energy interaction process between the electric vehicle and the charging station and the energy consumption during the journey to the charging station, so as to ensure the global optimality of charging decisions;

[0010] Step 4: Establish a multi-level goal programming model and construct a decision hierarchy structure;

[0011] Step 5: With the goal of selecting the optimal charging station, and based on the criteria of maximizing the energy requested by electric vehicles, minimizing the total response time, and minimizing charging costs, all charging stations are listed as alternatives. The priority weights of each criterion and alternative are compared and determined to minimize the weighted total deviation of each objective.

[0012] In step 1, the queuing system at each charging station is simulated using queuing theory. The user charging times are independent and follow an exponential distribution with parameter μ, whose probability density function is:

[0013] f(t) = μe -μt ,t≥0

[0014] Where: μ is the service rate (the average number of vehicles that complete service per unit time), the average service time is 1 / μ; t is the service time.

[0015] The time interval between user arrivals follows a Poisson distribution with parameter λ, and its probability density function is:

[0016] f(τ)=λe -λτ ,τ≥0

[0017] Where: λ is the arrival rate (the average number of vehicles arriving per unit time), the average arrival interval is 1 / λ; τ is the arrival interval.

[0018] y s (t+1)=y s (t)+e s (t)-l s (t) (1);

[0019] In equation (1): y s (t+1) represents the number of electric vehicles at the charging station at the next time step; y s (t) represents the number of electric vehicles currently at the charging station; e s (t) represents the number of electric vehicles currently entering the charging station; s(t) represents the number of electric vehicles currently leaving the charging station.

[0020]

[0021] In the formula: c s Indicates the number of service counters within the charging station; ρ s =λ s / μ s For system utilization; λ s Average arrival rate of electric vehicles at charging station s; μ s The average service rate of a single charging pile at charging station s; This represents the unnormalized state weights when the system has n electric vehicles. This represents the unnormalized state weights when there are exactly c (i.e., all charging stations are occupied, but there is no queue) vehicles in the system. This represents the decay rate of the system state probability for each additional vehicle added to the queue after the system enters a fully busy queuing state; n represents the number of electric vehicles in the system; n! is used to calculate the state probability weights when there are n electric vehicles in the system; c s ! is used to standardize the probability of the system state when all servers are occupied; P0 represents the probability of the system being idle, P n This represents the probability that the total number of tasks in the system is n at any given time, where all service desks are in a busy state.

[0022] Construct an M / M / c queuing system. Under steady state, e s =l s =λ s , where e s This indicates the number of electric vehicles entering the charging station; s This indicates the number of electric vehicles that have left the charging station.

[0023]

[0024] In equation (4): y s This indicates the number of electric vehicles at the charging station; Used to calculate the relative probability of a system being fully busy while there is still one vehicle waiting in the queue; ρ represents system utilization.

[0025] The probability that the service desk is busy is Average waiting time is

[0026]

[0027] In equation (5): This represents the probability that the system is completely busy.

[0028]

[0029] In step 1, the electric vehicle charging demand response model includes:

[0030] Simulated vehicle dynamics at charging stations: Equation (1);

[0031] Describing the randomness of user arrival and service: Equations (2) and (3);

[0032] Calculate the system congestion status: Equations (4) and (5), and quantify the user waiting time cost: Equation (6);

[0033] This provides a basis for the "response time" criterion in subsequent multi-objective optimization.

[0034] In step 2, to quantify the impact of traffic congestion on travel time, the BPR function is used to calculate the travel time T of a road segment. a (f);

[0035]

[0036] In equation (7): c ij (f) represents the time (travel cost) required to traverse the route; i, j represent the start and end points of the route; T a (f) represents the travel time of the road segment; f is the traffic volume of the road segment; M is the basic capacity of the road segment; T f The free-flow travel time for the road segment is denoted by α and β, which are constants, typically α = 0.15 and β = 4.

[0037] From the perspective of the overall road network, to achieve the goal of minimizing the total travel cost, a road minimum cost model is constructed, with the total travel cost being C. MC :

[0038]

[0039] In equation (8): f ij This represents the traffic volume on road segments i and j.

[0040] To quantify the contribution of travelers to traffic congestion, we calculate the additional travel cost C borne by a vehicle traveling between any two nodes i and j in the road network. E (i,j) represents the total extra travel time incurred by vehicles due to traffic congestion:

[0041]

[0042] In equation (9): N is the total number of vehicles in the road network that start at nodes i and j, respectively; L n f is the total number of road segments traversed by the l-th vehicle when traveling along the shortest path; lM represents the traffic volume of the l-th road segment; l This represents the traffic capacity of the l-th road segment; Let be the time required to travel through the l-th road segment under free-flow conditions. If the additional travel costs for travelers are significant, it indicates that vehicles are using a large number of congested road segments, and these travelers can be considered to be contributing significantly to the congestion of the traffic network. α represents the congestion delay multiplier, which represents the increase in travel time when the traffic flow f approaches the capacity M; β represents the congestion delay exponent, used to control the non-linear growth rate of the congestion effect with increasing flow.

[0043] Urban traffic dynamic congestion models specifically include:

[0044] BPR route travel time function: Equation (7);

[0045] The total travel cost model of the road network (8) and the additional travel cost model of individuals (9);

[0046] Together, they achieved the quantification of congestion from the macro road network to the micro individual level.

[0047] In step 3, the energy that needs to be replenished is calculated based on the current state of charge (SOC) and rated capacity of the electric vehicle (EV).

[0048]

[0049] In formula (10): This indicates the amount of energy requested by electric vehicle i; This represents the maximum state of charge of electric vehicle i, used to assess the available capacity of charging stations to ensure that charging stations retain a threshold of charge. Under the premise of rationally allocating the remaining energy; This indicates the current state of charge of electric vehicle i; This indicates the rated battery capacity of electric vehicle i.

[0050] Available energy at charging stations Represented as:

[0051]

[0052] In equation (11): This indicates the rated battery capacity of charging station s;

[0053] After charging, the energy status of the electric vehicle and the charging station is updated as follows:

[0054]

[0055] In the formula: This indicates the updated energy level of the electric vehicle i. This represents the current energy level of electric vehicle i; This indicates the amount of energy that charging station S provides to electric vehicle i; This represents the energy consumption of electric vehicle i during its journey to charging station S. This indicates the updated state of charge of charging station S; This represents the total amount of electricity provided by charging station S to all electric vehicles; I represents the electric vehicle number, used to iterate through all electric vehicles charging from this charging station.

[0056] Among these, the charging needs of electric vehicles must meet battery capacity limitations. Prevent battery overload. The energy provided by the charging station must not exceed its available capacity. To avoid affecting the stability of the power grid due to excessive discharge.

[0057] The energy consumption of electric vehicles (EVs) during driving is calculated using a longitudinal dynamics model:

[0058]

[0059] In equation (14): F(Se) i This represents the longitudinal traction force of an electric vehicle during operation, which is affected by air resistance, rolling resistance, slope resistance, and acceleration-related inertial forces. ω represents air density; ω is the drag coefficient; A is the frontal projected area of ​​the electric vehicle; Se i The average speed of electric vehicle i is represented by δ; the rolling coefficient is M. i The mass of electric vehicle i is represented by α; the road slope angle is represented by a. i For the acceleration of electric vehicles; To account for the equivalent mass of the inertial mass element; g is the gravitational acceleration.

[0060] The total energy cost of electric vehicle (EV) travel is:

[0061]

[0062]

[0063] In the formula: Let θ represent the total energy cost of the electric vehicle i traveling from its current location to the charging station Cs; θ represents the power transmission efficiency, taken as 0.8; D i→s This represents the straight-line distance from electric vehicle i to charging station Cs; This indicates the energy cost of the auxiliary load; Indicates energy consumption per unit distance; This represents the energy cost consumed by electric vehicle i to travel from its current location to charging station s.

[0064] Step 4 involves establishing a multi-level goal programming model, including:

[0065] 4.1: Objective function:

[0066]

[0067] In equation (17); P w P represents the objective of the hierarchical analysis weights; gu Indicates the energy supply target; P c Indicates the target capacity of charging stations; P T Indicates the response time target; P v Indicates the target charging cost; This indicates the bias of the weights in the hierarchical analysis. This indicates a deviation from the target energy supply target; This indicates that the charging station's capacity has exceeded the limit. This indicates that the response time has exceeded the limit. This indicates the user's budget overrun deviation.

[0068] 4.2: The constraints are as follows:

[0069]

[0070] In equation (18): I represents the electric vehicle number, used to iterate through all electric vehicles charging from this charging station; W i X represents the AHP weight of electric vehicle i to charging station s; i,s Let X be a binary decision variable. If electric vehicle i is assigned to charging station s, then X... i,s =1; otherwise 0; and , where represents the negative deviation (target not achieved) and positive deviation (target exceeded) of the AHP weight target, respectively; i represents the current electric vehicle number.

[0071]

[0072] In equation (19): This represents the updated energy of electric vehicle i; This represents the current energy of electric vehicle i; This represents the energy provided by charging station S to electric vehicle i; This represents the energy consumption of electric vehicle i during its journey to charging station S.

[0073]

[0074] In formula (20): I represents the number of the electric vehicle; C s This indicates the charging station with the number s; s is used to iterate through all available charging stations.

[0075]

[0076] In equation (21): This represents the total response time of electric vehicle i at charging station s; This represents the travel time of electric vehicle i to charging station s; This represents the waiting time of electric vehicle i at charging station s; This represents the charging time of electric vehicle i at charging station s; This represents the target response time value for electric vehicle i.

[0077]

[0078] In equation (22): p s V represents the electricity price charged by the charging station for electric vehicle i; i This represents the user's budget for electric vehicle i.

[0079] Constructing a decision hierarchy decomposes the multi-objective charging scheduling problem into three levels: objective, criterion, and solution.

[0080] The target layer represents the overall goal of scheduling and is used to select the optimal charging station.

[0081] The criteria layer decomposes the overall objective into multiple evaluation criteria through constraint equations (15) to (19), which are used to measure the merits of each candidate charging station.

[0082] The bottom layer lists all feasible alternative charging stations as options for decision-making. Each charging station will be scored according to the above criteria, and the optimal choice will be determined by a weighted composite score.

[0083] In step 5, the weight equation (18), constraint equations (19) to (22) and objective function equation (17) of the hierarchical analysis are encoded into a mathematical model. By transforming the unstructured charging scheduling problem into a mixed integer linear programming problem, the CPLEX solver is used to solve the problem, thereby outputting the optimal charging station allocation scheme that minimizes the overall deviation under all constraints.

[0084] In step 5, the goal is to select the optimal charging station, and the criteria are to maximize the energy requested by the electric vehicle (10), minimize the total response time (21), and minimize the charging cost (22). These three criteria are the core indicators for evaluating the merits of the candidate charging stations, and they are converted into deviation variables in the objective function for quantification.

[0085] All charging stations are listed as candidate options. The priority weights of each criterion and candidate option are determined by comparison. Using the AHP pairwise comparison matrix, the user or system decision-maker compares the relative importance of each criterion and calculates the weight vector of the criterion. For each criterion, the relative performance of the candidate charging stations is compared. Equation (18) obtains the final comprehensive weight by combining the criterion weights and the local weights of the options.

[0086] Equation (17) defines the objective of minimizing the weighted total deviation, transforming the multi-objective optimization problem into a single-objective optimization problem through the deviation variable. The MILP problem is solved using solvers such as CPLEX, outputting the optimal charging station allocation scheme that minimizes the weighted total deviation while satisfying all constraints. Finally, electric vehicle i is allocated to the charging station s that minimizes the overall deviation.

[0087] This invention provides a dynamic congestion-based charging scheduling method for electric vehicles based on multi-level objective programming, with the following technical advantages:

[0088] 1) Step 1 of this invention adopts the M / M / c queuing model, which can dynamically simulate the arrival and service process of electric vehicles in the charging station, establish a quantitative relationship between waiting time and user experience, adapt to the differentiated needs of traffic peak and off-peak periods, realize the precise spatiotemporal matching of charging demand, and improve the real-time and adaptability of scheduling decisions.

[0089] 2) Step 2 of this invention establishes an urban traffic network and a dynamic congestion model for charging and discharging to assess the impact of traffic congestion on charging behavior, transforming the degree of traffic congestion into a quantifiable travel cost indicator, enabling charging scheduling decisions to reflect changes in traffic conditions in real time, achieving a reasonable spatial distribution of charging demand, and improving the stability of the power distribution network and the carrying capacity of electric vehicles.

[0090] 3) Step 3 of this invention establishes an electric vehicle charging and discharging model, accurately quantifies the energy interaction process between the electric vehicle and the charging station based on the user's charging and discharging behavior characteristics, and considers the dynamic model to achieve a scientific assessment of the energy cost throughout the process, thereby realizing precise closed-loop management of energy interaction.

[0091] 4) Step 4 of this invention establishes a multi-level goal planning model, constructs a three-level decision-making hierarchy of "goal-criteria-solution", systematically coordinates and quantifies conflicting goals, determines the priority weight of each evaluation criterion, and uses the design of combined deviation variables to realize the transformation from qualitative preference to quantitative decision-making.

[0092] 5) Step 5 of this invention transforms the unstructured charging scheduling problem into a mixed integer linear programming problem and uses the CPLEX solver to achieve efficient solution, realizing dynamic balance optimization with multiple objectives, smoothing out load fluctuations of electric vehicles, reducing peak-valley differences in the distribution network, and improving the operating efficiency of the power grid.

[0093] 6) The charging scheduling method of the present invention dynamically adjusts the target weights and allocation strategies through multi-objective dynamic optimization and weight adaptation, adapts to the real-time changes in traffic, power grid and user demand, realizes personalized decision-making to meet diverse needs, improves the stability of the distribution network and increases the carrying capacity of the distribution network for electric vehicles. Attached Figure Description

[0094] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0095] Figure 1 This is a flowchart of an electric vehicle charging scheduling method based on multi-level objective programming.

[0096] Figure 2 This is a topology diagram of the transportation network.

[0097] Figure 3 This is a diagram showing the changes in charging vehicles within the charging station network.

[0098] Figure 4 This is a graph showing the load fluctuation of electric vehicles.

[0099] Figure 5 This is a graph showing the changes in distribution network losses. Detailed Implementation

[0100] like Figure 1 As shown, the dynamic congestion charging scheduling method for electric vehicles based on multi-level objective programming includes the following steps:

[0101] Step 1: Establish an electric vehicle charging demand response model based on the characteristics of electric vehicle user behavior, driving time, and electric vehicle charging waiting time costs;

[0102] Step 2: Based on the traffic flow dynamics and user charging / discharging density in different urban areas, establish an urban traffic dynamic congestion model that integrates the traffic network with the dynamic congestion of charging / discharging, thereby achieving accurate quantification of road travel time;

[0103] Step 3: Based on the user's charging and discharging behavior characteristics, establish an electric vehicle charging and discharging model to quantify the energy interaction process between the electric vehicle and the charging station and the energy consumption during the journey to the charging station, so as to ensure the global optimality of charging decisions;

[0104] Step 4: Establish a multi-level goal programming model and construct a decision hierarchy structure;

[0105] Step 5: With the goal of selecting the optimal charging station, and based on the criteria of maximizing the energy requested by electric vehicles, minimizing the total response time, and minimizing charging costs, all charging stations are listed as alternatives. The priority weights of each criterion and alternative are compared and determined to minimize the weighted total deviation of each objective.

[0106] Figure 2 This is the topology of the transportation network. The network includes 11 sets of travel demands, 20 links, and 12 nodes, with a free-roaming speed of 80 km / h. The vehicle transmission efficiency is 0.8, and the frontal projected area is 2.27 m². 2 The rolling resistance coefficient is 0.01, and the air density is 1.293 kg / m³. 3 The slope angle is as follows: During the off-peak period, the number of vehicles passing through is reduced to 0.19 of the road's carrying capacity; during the mid-off-peak period, the number of vehicles passing through is reduced to 0.475 of the road's carrying capacity; during the mid-peak period, the number of vehicles passing through is reduced to 0.713 of the road's carrying capacity; and during the peak period, the number of vehicles passing through is the full road carrying capacity.

[0107] Figure 3 This diagram illustrates the changes in charging vehicles within the charging station network. It employs a 33-node IEEE distribution network coupled with a 12-node traffic network system. The total load of the distribution network is 3715kW + 2300kVar, the base voltage is 12.66kV, and the base power is 10MW. The simulation includes 1200 electric vehicles: 800 private cars, 200 taxis, and 200 buses. Each charging pile group can charge 30 vehicles. The battery capacity for private cars and taxis is 30kWh, with a maximum charge / discharge power of 30kW. The battery capacity for buses is 150kWh, with a maximum charge / discharge power of 75kW. The charge / discharge efficiency for all three types of electric vehicles is 0.9, and the charging threshold for electric vehicles is 0.2. Three electric vehicle charging stations are set up in the transportation network, with locations and capacities of {6, 8, 11} and {600, 400, 800} kW respectively. The electricity price is RMB 0.97 / kWh in the peak period (9:00-11:00, 15:00-21:00), RMB 0.61 / kWh in the parity period (7:00-8:00, 11:00-12:00, 13:00-14:00, 22:00-23:00), and RMB 0.35 / kWh in the low-price period (0:00-7:00, 12:00-13:00). The maximum charging and discharging power is 250 kW, the discharge efficiency is 0.9, and the simulation time step is 1 min.

[0108] Figure 4 and Figure 5 The figures show the load fluctuations of electric vehicles and the changes in power distribution network losses, respectively. Figure 4 The load changes of electric vehicles are more stable, the peak-to-valley difference is smaller, and the overall distribution network loss is reduced. Experiments show that by meeting the charging demand expectations of electric vehicle users and adjusting the weighted deviation of each objective, EV load fluctuations and distribution network losses can be reduced, further reducing the risk of voltage exceeding limits.

Claims

1. A dynamic congestion charging scheduling method for electric vehicles based on multi-level objective programming, characterized in that... Includes the following steps: Step 1: Establish an electric vehicle charging demand response model based on the characteristics of electric vehicle user behavior, driving time, and electric vehicle charging waiting time costs; Step 2: Based on the traffic flow dynamics and user charging / discharging density in different urban areas, establish an urban traffic dynamic congestion model that integrates the traffic network with the charging / discharging dynamic congestion degree. Step 3: Based on the user's charging and discharging behavior characteristics, establish an electric vehicle charging and discharging model to quantify the energy interaction process between the electric vehicle and the charging station and the energy consumption during the journey to the charging station; Step 4: Establish a multi-level goal programming model; Step 5: With the goal of selecting the optimal charging station, and based on the criteria of maximizing the energy requested by electric vehicles, minimizing the total response time, and minimizing charging costs, all charging stations are listed as alternatives. The priority weights of each criterion and alternative are compared and determined to minimize the weighted total deviation of each objective.

2. The electric vehicle dynamic congestion charging scheduling method based on multi-level objective programming according to claim 1, characterized in that: In step 1, the queuing system at each charging station is simulated using queuing theory. The user charging times are independent and follow an exponential distribution with parameter μ, whose probability density function is: f(t)=μe -μt ,t≥0 Where: μ is the service rate, the average service time is 1 / μ; t is the service time.

3. The electric vehicle dynamic congestion charging scheduling method based on multi-level objective programming according to claim 2, characterized in that: The time interval between user arrivals follows a Poisson distribution with parameter λ, and its probability density function is: f(τ)=λe -λτ ,τ≥0 Where: λ is the arrival rate, the average arrival interval is 1 / λ; τ is the arrival interval; y s (t+1)=y s (t)+e s (t)-l s (t) (1); In equation (1): y s (t+1) represents the number of electric vehicles at the charging station at the next time step; y s (t) represents the number of electric vehicles currently at the charging station; e s (t) represents the number of electric vehicles currently entering the charging station; s (t) represents the number of electric vehicles currently leaving the charging station.

4. The electric vehicle dynamic congestion charging scheduling method based on multi-level objective programming according to claim 3, characterized in that: In the formula: c s Indicates the number of service counters within the charging station; ρ s =λ s / μ s For system utilization; λ s Average arrival rate of electric vehicles at charging station s; μ s The average service rate of a single charging pile at charging station s; This represents the unnormalized state weights when the system has n electric vehicles. This represents the unnormalized state weights when there are exactly c vehicles in the system. This represents the decay rate of the system state probability for each additional vehicle added to the queue after the system enters a fully busy queuing state; n represents the number of electric vehicles in the system; n! is used to calculate the state probability weights when there are n electric vehicles in the system; c s ! is used to standardize the probability of the system state when all servers are occupied; P0 represents the probability of the system being idle; P n This represents the probability that the total number of tasks in the system is n at any given time.

5. The electric vehicle dynamic congestion charging scheduling method based on multi-level objective programming according to claim 4, characterized in that: Construct an M / M / c queuing system. Under steady state, e s =l s =λ s , where e s This indicates the number of electric vehicles entering the charging station; s This indicates the number of electric vehicles that have left the charging station; In equation (4): y s This indicates the number of electric vehicles at the charging station; Used to calculate the relative probability of a system being fully busy while there is still one vehicle waiting in the queue; ρ represents system utilization. The probability that the service desk is busy is Average waiting time is In equation (5): Indicates the probability that the system is completely busy; 6. The electric vehicle dynamic congestion charging scheduling method based on multi-level objective programming according to claim 5, characterized in that: In step 2, to quantify the impact of traffic congestion on travel time, the BPR function is used to calculate the travel time T of a road segment. a (f); In equation (7): c ij (f) represents the time required to traverse the road segment; i and j represent the start and end points of the road segment. T a (f) indicates the travel time for the route segment; f represents the traffic volume of the road segment; M represents the basic traffic capacity of the road segment; T f The free-flow travel time for the road segment; α and β are constants; From the perspective of the overall road network, to achieve the goal of minimizing the total travel cost, a road minimum cost model is constructed, with the total travel cost being C. MC : In equation (8): f ij Indicates the traffic volume on road segments i and j; To quantify the contribution of travelers to traffic congestion, we calculate the additional travel cost C borne by a vehicle traveling between any two nodes i and j in the road network. E (i,j) represents the total extra travel time incurred by vehicles due to traffic congestion: In equation (9): N is the total number of vehicles in the road network that start at nodes i and j, respectively; L n f is the total number of road segments traversed by the l-th vehicle when traveling along the shortest path; l Let L be the traffic volume of the l-th road segment; M l This represents the traffic capacity of the l-th road segment; α represents the time required to pass through the l-th road segment under free-flow conditions; if the additional travel costs for travelers are high, it indicates that vehicles use a large number of congested road segments, and these travelers are considered to contribute significantly to the congestion of the traffic network; α represents the congestion delay multiplier coefficient; β represents the congestion delay index.

7. The method for dynamic congestion charging scheduling of electric vehicles based on multi-level objective programming according to claim 6, characterized in that: In step 3, the energy that needs to be replenished is calculated based on the current state of charge (SOC) and rated capacity of the electric vehicle (EV). In formula (10): This indicates the amount of energy requested by electric vehicle i; This represents the maximum state of charge of electric vehicle i, used to assess the available capacity of charging stations to ensure that charging stations retain a threshold of charge. Under the premise of rationally allocating the remaining energy; This indicates the current state of charge of electric vehicle i; This indicates the rated battery capacity of electric vehicle i; Available energy at charging stations Represented as: In equation (11): This indicates the rated battery capacity of charging station s; After charging, the energy status of the electric vehicle and the charging station is updated as follows: In the formula: This indicates the updated energy level of the electric vehicle i. This represents the current energy level of electric vehicle i; This indicates the amount of energy that charging station S provides to electric vehicle i; This represents the energy consumption of electric vehicle i during its journey to charging station S. This indicates the updated state of charge of charging station S; This represents the total amount of electricity provided by charging station S to all electric vehicles; I represents the number of the electric vehicle; and the charging demand of the electric vehicles must meet the battery capacity limit. To prevent battery overload, the energy provided by the charging station must not exceed its available capacity. To avoid affecting the stability of the power grid due to excessive discharge.

8. The electric vehicle dynamic congestion charging scheduling method based on multi-level objective programming according to claim 7, characterized in that: The energy consumption of electric vehicles (EVs) during driving is calculated using a longitudinal dynamics model: In equation (14): F(Se) i This represents the longitudinal traction force of an electric vehicle during operation, which is affected by air resistance, rolling resistance, slope resistance, and acceleration-related inertial forces. ω represents air density; ω is the drag coefficient; A is the frontal projected area of ​​the electric vehicle; Se i The average speed of electric vehicle i is represented by δ; the rolling coefficient is M. i The mass of electric vehicle i is represented by α; the road slope angle is represented by a. i For the acceleration of electric vehicles; To account for the equivalent mass of the inertial mass element; g is the acceleration due to gravity; The total energy cost of electric vehicle (EV) travel is: In the formula: θ represents the total energy cost of electric vehicle i traveling from its current location to charging station Cs; θ represents the powertrain efficiency; D i→s This represents the straight-line distance from electric vehicle i to charging station Cs; This indicates the energy cost of the auxiliary load; Indicates energy consumption per unit distance; This represents the energy cost consumed by electric vehicle i to travel from its current location to charging station s.

9. The method for dynamic congestion charging scheduling of electric vehicles based on multi-level objective programming according to claim 8, characterized in that: Step 4 involves establishing a multi-level goal programming model, including: 4.1: Objective Function: In equation (17); P w P represents the objective of the hierarchical analysis weights; gu Indicates the energy supply target; P c Indicates the target capacity of charging stations; P T Indicates the response time target; P v Indicates the target charging cost; This indicates the bias of the weights in the hierarchical analysis. This indicates a deviation from the target energy supply target; This indicates that the charging station's capacity has exceeded the limit. This indicates that the response time has exceeded the limit. This indicates the user's budget overrun deviation; 4.2: The constraints are as follows: In equation (18): I represents the electric vehicle number, used to iterate through all electric vehicles charging from this charging station; W i X represents the AHP weight of electric vehicle i to charging station s; i,s Let X be a binary decision variable. If electric vehicle i is assigned to charging station s, then X... i,s =1; otherwise 0; and , where represents the negative and positive deviations of the AHP weight target, respectively; i represents the current electric vehicle number; In equation (19): This represents the updated energy of electric vehicle i; This represents the current energy of electric vehicle i; This represents the energy provided by charging station S to electric vehicle i; This represents the energy consumption of electric vehicle i during its journey to charging station S. In formula (20): I represents the number of the electric vehicle; C s This represents the charging station with number s; s is used to iterate through all available charging stations. In equation (21): This represents the total response time of electric vehicle i at charging station s; This represents the travel time of electric vehicle i to charging station s; This represents the waiting time of electric vehicle i at charging station s; T represents the charging time of electric vehicle i at charging station s; i d This represents the target response time value for electric vehicle i; In equation (22): p s V represents the electricity price charged by the charging station for electric vehicle i; i This represents the user's budget for electric vehicle i.

10. The method for dynamic congestion charging scheduling of electric vehicles based on multi-level objective programming according to claim 9, characterized in that: In step 5, the goal is to select the optimal charging station, and the criteria are to maximize the energy requested by the electric vehicle (10), minimize the total response time (21), and minimize the charging cost (22). All charging stations are listed as alternatives. The priority weights of each criterion and alternative are determined by comparison. The relative importance of each criterion is compared by the pairwise comparison matrix of AHP, and the weight vector of the criterion is calculated. For each criterion, the relative performance of the candidate charging stations is compared; Equation (18) obtains the final comprehensive weight by combining the criterion weights and the local weights of the schemes; Equation (17) defines the objective of minimizing the weighted sum of deviations, and transforms the multi-objective optimization problem into a single-objective optimization problem through the deviation variable; The MILP problem is solved using solvers such as CPLEX, and the optimal charging station allocation scheme that minimizes the total weighted deviation under all constraints is output. Finally, electric vehicle i is assigned to the charging station s that minimizes the overall deviation.

Images