Modeling method for electric vehicle charging behavior portrait

Abstract

The invention provides a modeling method for a charging behavior portrait of an electric vehicle user. The modeling method is specifically implemented according to the following steps: step 1, obtaining sample data of the electric vehicle charging user by using a Monte Carlo sampling method; step 2, establishing a user mileage anxiety model; step 3, establishing a queuing model of the electric vehicle users; and step 4, establishing a behavior portrait of the electric vehicle user. The arrival time, daily mileage and mileage anxiety psychology of the electric vehicle user and the charging decision of the electric vehicle user are obtained by adopting multiple methods such as Monte Carlo sampling, a mileage anxiety model and a queuing model of the electric vehicle user, and then the behavior portrait of the electric vehicle charging user is constructed.

Description

Technical Field

[0001] The present invention belongs to the field of electric vehicle charging, and in particular is a modeling method for an electric vehicle charging behavior portrait. Background Art

[0002] In recent years, with the popularization of electric vehicles, electric vehicles have developed rapidly, and the number of electric vehicles has continued to rise. Due to the rapid popularization of electric vehicles and the lack of charging infrastructure, electric vehicles have serious problems such as low charging efficiency, poor user experience and waste of resources. In order to solve the above problems, charging station operators need to optimize the layout and management of charging stations according to the charging behavior of electric vehicle users, so as to improve the quality and efficiency of charging services and the operating benefits of charging stations. Therefore, it is urgent to establish a modeling method for the charging behavior portrait of electric vehicle users and accurately construct the charging behavior portrait of electric vehicle users. This work can provide a theoretical basis for charging demand response to guide users to charge in an orderly manner, and also provide a basis for charging load prediction of charging stations. Summary of the invention

[0003] In order to solve the above technical problems, the purpose of the present invention is to provide a method for constructing a charging behavior profile of electric vehicle users, which constructs an accurate charging user behavior profile of electric vehicles through the arrival time, charging time, daily mileage, mileage anxiety and the probability of charging at a charging station of electric vehicle users. The problem that the processing of charging behavior of charging users in the prior art does not take into account the user's mileage anxiety, and if the queue at the power station is long, the electric vehicle user may become impatient and leave, thereby leading to inaccurate estimation of charging demand of the charging station.

[0004] In order to achieve the above-mentioned purpose of the present invention, the present invention adopts the following technical scheme:

[0005] A modeling method for a charging behavior portrait of an electric vehicle user comprises the following steps:

[0006] Step 1: Statistical analysis is performed on the historical charging users of the charging station, and the sample data of electric vehicle charging users is obtained using the Monte Carlo sampling method, including the arrival time of the electric vehicle, the SOC during charging, and the daily mileage;

[0007] Step 2: Establish a user mileage anxiety model and obtain the mileage anxiety degree K s ;

[0008] Step 3: Establish a queuing model for electric vehicle users to charge;

[0009] Step 4: Build a behavioral profile of electric vehicle users;

[0010] Wherein, step 1 is specifically implemented according to the following steps:

[0011] Step 1.1: Perform statistical analysis on the historical charging users of the charging station and establish a probability distribution model for the initial charging time, daily mileage, charging time, and initial state of charge of electric vehicles. The specific contents are as follows:

[0012] 1) The initial charging time of electric vehicle users generally conforms to the normal distribution, and its probability density function is:

[0013]

[0014] Where t represents the initial charging time, μ a , σ a are the mean and standard deviation of their distributions, respectively.

[0015] 2) The daily mileage of electric vehicles follows a log-normal distribution, and its probability density function is

[0016]

[0017] Where x is the mileage in kilometers. b and σ b are their expectation and standard deviation respectively.

[0018] 3) The battery state SOC of the electric vehicle when charging. The battery state calculation formula is:

[0019] S OC =(S OC0 -x / x max )×10000

[0020] S OC represents the battery status of the electric vehicle when the electric vehicle user arrives at the charging station, S OC0 It represents the battery status of the electric vehicle before the electric vehicle user travels, and x is the mileage.

[0021] Step 1.2: Based on the probability density functions of the above electric vehicles, the Monte Carlo method is used to generate electric vehicle charging sample data to obtain the arrival time, SOC during charging, and daily mileage of the electric vehicle;

[0022] Step 2 is implemented as follows:

[0023] Quantitative modeling of the range anxiety of different electric vehicle users.

[0024] Step 2.1: Establish the relationship between the state of charge of the electric vehicle at the end of charging on the day and the travel mileage on the second day. The specific relationship model is as follows:

[0025]

[0026]

[0027] Among them, Q(S oc ,x) is when the travel mileage is x and the SOC level is S oc The stopping probability density function at ; is the stopping probability density function when SOC cannot meet the travel mileage demand, SOC is S oc The mileage that an electric vehicle can travel, w is the energy consumption per unit mileage of the electric vehicle, S min is the minimum SOC of the electric vehicle power battery, and E0 is the initial capacity of the electric vehicle power battery, in kW·h.

[0028] Step 2.2: Considering the daily mileage of electric vehicles and the state of charge during charging, a quantitative model of mileage anxiety for electric vehicle users is established. The specific contents are as follows:

[0029] If the SOC of an electric vehicle is low at the time of travel, the psychological anxiety of the electric vehicle user will obviously be high during the travel, and vice versa. Therefore, the concern of the electric vehicle user about not being able to complete the travel is quantified as the mileage anxiety K based on the daily mileage of the electric vehicle and the state of charge of the electric vehicle. s , the formula is as follows:

[0030]

[0031] In the formula, Q s is the probability of electric vehicle users stopping driving, Q smin is the minimum stopping probability of electric vehicles; Q smax is the maximum stopping probability of electric vehicles; L Smax is the mileage that the car can travel when the battery is fully charged; ε is the travel mileage greater than L smax probability.

[0032] Step 3: Establish the electric vehicle multi-subprocess admission control scheme and the user charging queuing model.

[0033] Step 3.1: First establish M / T based D / n / (n+b) electric vehicle multi-subprocess admission control scheme, such a multi-subprocess admission scheme can be modeled as a queue with 0 buffer (b=0) and n subprocesses (M / T D / n / n).

[0034] Furthermore, in the electric vehicle admission control scheme, M represents the number of electric vehicles arriving, T Dis the time a subprocess is occupied; n is the number of subprocesses in the admission process; b is the buffer of the admission process, and its size is zero. When the buffer is zero, if the EV arrives and finds that all charging piles are busy (that is, all subprocesses are occupied), the EV will be denied service.

[0035] Furthermore, under the admission control scheme, if all charging piles are occupied, electric vehicles will not be allowed to enter. Otherwise, electric vehicles will be charged during the charging service time T D Occupies idle charging piles.

[0036] After being accepted by the admission control scheme, EVs will wait in the charging queue and receive charging services at the charging station according to the first-in-first-out (FIFO) strategy.

[0037] Step 3.2: Further, a queuing model based on M / M2 / m is established on the basis of admission control, where M represents the average number of electric vehicles arriving; M2 represents the charging power of the charging pile; and m represents the number of charging piles.

[0038] After passing the admission control, electric vehicles allowed to enter the charging station will wait in the queue. The expression of the electric vehicle charging user queue model is as follows:

[0039]

[0040]

[0041] Among them, l t is the queue length of the charging station at time t; is the average number of electric vehicles arriving at time t; m is the total number of electric vehicle charging piles at the charging station; Pr 0,t is the probability that all charging piles in the charging station are idle at time t; k is the number of idle charging piles in the charging station; D t is the total charging demand of the charging station at time t; d is the average charging demand of each electric vehicle, and μ is the fixed charging power of each charging pile.

[0042] Step 4: Create a behavioral profile of electric vehicle users queuing at charging stations. The specific content is as follows:

[0043] The behaviors of electric vehicle users after arriving at the charging station are divided into two types: refusing to join the queue and leaving the charging station impatiently;

[0044] Furthermore, when the electric car user arrives at the charging station, he finds that the queue is too long and therefore refuses to join the queue. The probability model of the electric car user refusing to join the charging service provided by the charging station is:

[0045]

[0046] Among them, Pr t (RTJ|adm) is the conditional probability that an electric vehicle user refuses to join the charging station queue; Pr t RTJ Pr is the probability that the user refuses to join the charging station queue; t (adm|RTJ) is the conditional probability that the user who refuses to charge passes the admission control scheme; Pr t adm is the probability that the user passes the admission control scheme; σ is the parameter that defines the probability decrease rate.

[0047] Furthermore, when electric vehicle users become impatient after waiting in line for too long, they will also choose to leave the queue. The probability that electric vehicle users become impatient and leave the charging station is:

[0048]

[0049] Among them, Pr t (IL|adm) is the conditional probability that the accepted user becomes impatient and leaves the charging station; Pr t IL is the probability that a user leaves the queue due to impatience; Pr(adm|IL) is the probability that a user leaves the charging station due to impatience through the admission control scheme; δ is a non-negative adjustment factor.

[0050] Furthermore,

[0051] Pr t (adm|RTJ)=1,Pr t (adm|IL)=1

[0052]

[0053] The present invention can obtain data on different aspects of electric vehicle users and perform portrait modeling on the charging behavior of electric vehicles arriving at charging stations. Specifically, a behavioral portrait of electric vehicle charging users is constructed based on the electric vehicle users' arrival time, charging time, daily mileage, mileage anxiety, and the probability of charging at the charging station.

[0054] The beneficial effects of the present invention are as follows: The present invention combines Monte Carlo sampling, mileage anxiety model, access control scheme and physical queuing model to provide an accurate modeling scheme for establishing a charging behavior portrait of electric vehicle users. First, based on the statistical analysis of historical data, a multi-faceted probability model of electric vehicles is established, and the charging data of electric vehicles is generated by Monte Carlo sampling to obtain the user's arrival time, charging time and daily mileage, which effectively captures the diversity and uncertainty of user behavior. Secondly, the mileage anxiety model takes into account the changes in the user's psychological state that may occur during the charging process, thereby more realistically describing the user behavior in the actual scene. Finally, the probability of the user charging is calculated through the access control scheme and the physical queuing model, and the utilization efficiency of the charging facilities can be improved and the waiting time of the user can be reduced. The access control scheme can reasonably allocate charging resources according to the user's needs and the actual situation of the charging facilities, so as to maximize the satisfaction of user needs while minimizing the idle time of the charging facilities. The physical queuing model is based on queuing theory, and by optimizing the operation mode of the charging facilities, it effectively reduces the time for users to wait for charging and improves the quality and efficiency of the overall charging service. The present invention provides important technical support and guidance for promoting the popularization of electric vehicles and the construction of charging infrastructure. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a charging behavior portrait of electric vehicle users established by the present invention;

[0056] Figure 2 The present invention is based on M / T D / n / (n+b) working flow diagram of electric vehicle access control scheme;

[0057] Figure 3 This is a physical queuing model diagram based on M / M2 / N of the present invention. DETAILED DESCRIPTION

[0058] The present invention is described in detail below with reference to the accompanying drawings.

[0059] like Figure 1 As shown, a method for modeling electric vehicle user behavior portrait is implemented in the following steps:

[0060] Step 1: Collect historical data of electric vehicle users at charging stations, establish a probability model of electric vehicle users, and obtain the user's arrival time, battery status of electric vehicles, and daily mileage;

[0061] Step 2: Establish a user range anxiety model to obtain the psychological state of electric vehicle users;

[0062] Step 3: Obtain the user's queueing time, the probability of refusing to queue, and the probability of leaving the queue due to waiting too long, and establish a M / T-based D / n / (n+b) electric vehicle admission control scheme and physical queuing model based on M / M2 / N;

[0063] Step 4: Build a behavioral profile of electric vehicle users;

[0064] like Figure 2 As shown, step 3 builds based on M / T D / n / (n+b) electric vehicle admission control scheme: In order to strike a balance between the admission rate and the utilization rate of charging piles, a multi-subprocess admission control scheme consisting of n subprocesses is adopted. When the same subprocess accepts two electric vehicles in succession, their inter-arrival time must be greater than the service time T of the charging pile. d , when electric vehicle users can fit into one of the sub-processes, they are allowed access.

[0065] Furthermore, the specific working process of the admission control scheme of the two sub-processes is as follows Figure 3 As shown:

[0066] An electric vehicle admission control scheme with two sub-processes. When electric vehicle user 1 arrives, it is assigned to sub-process 1. When electric vehicle charging user 2 arrives, since the arrival interval between electric vehicle user 1 and electric vehicle charging user 2 is less than the charging time T of the electric vehicle user, d (length of the shaded rectangle), subprocess 1 is occupied and user 2 cannot be assigned to subprocess 1. Therefore, EV user 2 is assigned to subprocess 2.

[0067] Furthermore, when electric car user 3 arrives, he is assigned to subprocess 1, which is not occupied at this time. When electric car user 4 arrives, since both subprocesses are occupied, no subprocess can be assigned. Therefore, the charging station will not accept electric car user 4.

[0068] Furthermore, when EV user 5 arrives, sub-process 2 is idle at this time, and it is then assigned to sub-process 2.

[0069] Furthermore, the admission control scheme that allows electric vehicles to access charging stations can ensure the stability of the queuing process of electric vehicle users and achieve a balance between waiting time, admission probability and charging pile utilization.

[0070] Furthermore, in order to ensure the stability of the user queuing process, the service time in the admission control process must be greater than the average charging time of the user, which is expressed as:

[0071] nTD >md / μ t

[0072] Among them, T D is the service time of a subprocess; n is the number of subprocesses in the admission control process; m is the total number of electric vehicle charging piles in the charging station; μ t is the fixed charging power for each charging pile; d is the charging demand of each electric vehicle. Further, the above inequality is converted into an equality constraint:

[0073]

[0074] in is a coefficient greater than 1. The determination of an admission control strategy involves two decision variables and n, these two decision variables can be used to calculate T d T D .

[0075] Furthermore, the probability of allowing electric vehicles to access charging stations can be expressed as:

[0076]

[0077] Among them, Pr t adm Probability of charging access for electric vehicles; is the arrival rate of electric vehicles at the charging station at time t; Γ(.) is the gamma function.

[0078] like Figure 3 As shown in the figure, after the electric vehicle user passes the admission control scheme, a charging queue model based on M / M2 / m is established. The specific steps are as follows:

[0079] After the admission control scheme, the admitted electric vehicles will wait in the queue. The M / M2 / m queue model is adopted in the queuing process, where M represents the arrival rate of electric vehicles; M2 represents the charging power of electric vehicles; and m represents the number of electric vehicle charging piles at the charging station.

[0080] Furthermore, the expression of the user queuing model is as follows:

[0081]

[0082] Among them, Pr 0,t is the probability that all charging piles in the charging station are idle at time t; k is the number of idle charging piles in the charging station.

[0083] Furthermore, the behavior of electric vehicle users occurs after the charging station’s access control. Figure 3 In Figure 2, a user charging queuing model based on the admission control scheme is shown.

[0084] Furthermore, after electric vehicle users arrive at the charging station, the users are divided into two types of user behaviors: those who refuse to join the queue and those who leave the charging station due to impatience caused by the long queue time;

[0085] One user behavior is that electric vehicle users refuse to join the queue. When electric vehicle users arrive at the charging station, they find that the queue is too long and refuse to join the queue. The probability model of electric vehicle users refusing to queue under admission control is:

[0086]

[0087]

[0088] Among them, σ is a parameter that defines the probability decrease rate.

[0089] Another user behavior is that electric car users leave the charging station because of long waiting time. The probability that electric car users leave the charging station due to impatience is:

[0090]

[0091] Pr t lL =δln(l t -m+1) / l t

[0092] Where δ is a non-negative adjustment factor.

[0093] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A modeling method for electric vehicle user charging behavior portrait, characterized in that: The steps include: 1) Conduct statistical analysis on the historical charging users of the charging station and use the Monte Carlo sampling method to obtain sample data of electric vehicle charging users; 2) Establish a user mileage anxiety model; 3) Establish a queuing model for electric vehicle users; 4) Establish behavioral profiles of electric vehicle users.

2. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 1 is characterized in that: In step 1), we first conduct a statistical analysis of the historical charging users of the charging station and establish a probability distribution model of the initial charging time, daily mileage, and initial state of charge of electric vehicle users. The specific contents are as follows: 1) The initial charging time of electric vehicle users generally conforms to the normal distribution, and its probability density function is: Where t represents the initial charging time, μ a , σ a are the expectation and standard deviation of its distribution respectively; 2) The daily mileage of electric vehicles follows a log-normal distribution, and its probability density function is Where x is the mileage, unit: kilometers, μ b and σ b are their expectation and standard deviation respectively; 3) The battery state SOC (State of charge) of an electric vehicle when it is charging is calculated as S OC =(S OC0 -x / x max )×100% In the formula, S OC Indicates the battery status of an electric vehicle user when he arrives at a charging station and prepares to charge. OC0 It represents the battery status of the electric vehicle before the electric vehicle user travels, and x is the mileage.

3. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 2 is characterized in that: According to the probability density functions of various aspects of electric vehicles, the Monte Carlo method is used to generate electric vehicle charging sample data.

4. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 1 is characterized in that In step 2), the daily mileage of the electric vehicle and the state of charge during charging are considered to quantitatively model the mileage anxiety of different electric vehicle users.

5. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 4 is characterized in that: In step 2), the relationship between the state of charge of the electric vehicle at the end of charging on the day and the travel mileage on the second day is first established. The specific relationship model is as follows: Among them, Q(S oc ,x) is when the travel mileage is x and the SOC level is S oc The stopping probability density function at ; is the stopping probability density function when SOC cannot meet the travel mileage demand, SOC is S oc The mileage that an electric vehicle can travel, w is the energy consumption per unit mileage of the electric vehicle, S min is the minimum SOC of the electric vehicle power battery, E0 is the electric vehicle power battery capacity, in kW·h.

6. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 5 is characterized in that: In step 2), a mileage anxiety quantification model for electric vehicle users is further established, and the mileage anxiety degree of electric vehicle users is quantified as mileage anxiety degree K s , the model formula is as follows: In the formula, Q s is the probability of electric vehicle users stopping driving, Q smin is the minimum stopping probability of electric vehicles; Q smax is the maximum stopping probability of electric vehicles; L smax is the mileage that the car can travel when the battery is fully charged; ε is the travel mileage greater than L smax probability. That is f(x).

7. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 1 is characterized in that In step 3), first establish the M / T D / n / (n+b) electric vehicle charging admission control scheme, where M represents the arrival rate of electric vehicles, T D is the time a subprocess is occupied (i.e., the deterministic service time of a charging pile); n represents the number of subprocesses in the admission process; b is the buffer of the admission process, which is set to zero in this scheme, that is, under the admission control scheme, if all subprocesses are occupied, electric vehicles will not be admitted, otherwise, electric vehicles will be allowed to enter the charging station and wait in line.

8. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 1 is characterized in that In step 3), a queue model based on M / M2 / m is further established, where M represents the arrival rate of electric vehicles; M2 represents the charging power; and N represents the number of charging piles. After passing the admission control, electric vehicles allowed to access the charging station will wait in the queue. The expression of the queue length of electric vehicle charging users is as follows: Among them, l t is the queue length of the charging station at time t; is the average number of electric vehicles arriving at time t; m is the total number of electric vehicle charging piles at the charging station; Pr 0,t is the probability that all charging piles in the charging station are idle at time t; k is the number of idle charging piles in the charging station; D t is the total charging demand of the charging station at time t; d is the charging demand of each electric vehicle, and μ is the fixed charging power of each charging pile.

9. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 1 is characterized in that: In step 4), the behavior of electric vehicle users after arriving at the charging station is modeled and divided into two types: refusing to join the queue and leaving the charging station impatiently. The probability model of electric vehicle users refusing to join the charging station to provide charging services under admission control is: Among them, Pr t (RTJ|adm) is the conditional probability that an electric vehicle user refuses to join the charging station queue; Pr is the probability that the user refuses to join the charging station queue; t (adm|RTJ) is the conditional probability that the user who refuses to charge passes the admission control scheme; is the probability of a user passing the admission control; σ is the parameter that defines the probability drop rate; The probability that an electric car user becomes impatient and leaves the charging station is: Among them, Pr t (IL|adm) is the conditional probability that the accepted user becomes impatient and leaves the charging station; is the probability that a user leaves the queue due to impatience; Pr(adm|IL) is the probability that a user leaves the charging station due to impatience through the admission control scheme; δ is a non-negative adjustment factor.

10. The modeling method of the charging behavior portrait of an electric vehicle user according to claim 1 is characterized in that: A behavioral profile of electric vehicle charging users is constructed based on their arrival time, charging time, daily mileage, mileage anxiety, and the probability of charging at a charging station.

Images