Electric vehicle charging load prediction method based on order data

Abstract

The invention discloses an electric vehicle charging load prediction method based on order data, and the method comprises the following steps: S1, carrying out the behavior analysis of an electric vehicle based on order data, the initial charging time probability distribution, the initial charging state of charge of the electric vehicle, the charging power level probability distribution and the charging duration of the electric vehicle in different environments are obtained; s2, according to the behavior analysis result of the step S1, performing random simulation by using a Monte Carlo method to obtain the daily driving mileage, the initial charging time and the initial battery SOC of the electric vehicle; and S3, according to the stochastic simulation result of the step S2, based on a charging load calculation formula, predicting to obtain a charging load of a physical level. According to the method, the charging load of the electric vehicle can be predicted and analyzed from the physical perspective based on the order data, the action relationship between the electricity price and the charging load is further analyzed from the economic level, and the charging load change of the electric vehicle under economic stimulation is predicted, so that the accuracy and generalization of load prediction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging load prediction, and in particular to an electric vehicle charging load prediction method based on order data. Background Art

[0002] As the penetration rate of electric vehicles continues to increase, their charging behavior has brought challenges to the stable operation of the power grid. Accurately predicting the charging demand of electric vehicles is of great significance for analyzing the impact of electric vehicles on the grid, power system planning and stable operation of the grid. However, the charging demand of electric vehicles is affected by many factors, and the charging load of electric vehicles has strong randomness and volatility, which makes it difficult to predict the charging demand.

[0003] In recent years, domestic and foreign scholars have done a lot of research on the prediction of electric vehicle charging load. The mainstream methods are mainly divided into two types: prediction methods based on short-term historical loads of charging stations and load prediction methods based on Monte Carlo simulation. The former method focuses on short-term and small time scale electric vehicle load prediction. This type of method generally uses machine learning and deep learning to establish a black box model and predicts the charging load of electric vehicles through model training. It is suitable for microscopic scenarios such as real-time power control of electric vehicles. However, this type of method is difficult to discover the essence of charging load changes and has poor generalization ability. The model needs to be retrained in different scenarios. The load prediction method based on Monte Carlo simulation is to discover the essence and law of load changes through the analysis of electric vehicle behavior characteristics and derive the probability distribution under different scenarios, and use the Monte Carlo simulation method to predict the charging load of electric vehicles. The load prediction method based on Monte Carlo simulation only considers conventional influencing factors such as the number of electric vehicles, daily mileage, and initial state of charge when modeling. However, the charging station is the key link in the transformation from behavior to load. The research on electric vehicle behavior should be strengthened while also considering the transformation process from charging station behavior to load. There is a lack of analysis of electric vehicle charging behavior at the economic level such as time-of-use electricity prices and charging subsidies. Summary of the invention

[0004] In view of the above problems, the present invention aims to provide a method for predicting electric vehicle charging load based on order data.

[0005] The technical solution of the present invention is as follows:

[0006] A method for predicting electric vehicle charging load based on order data comprises the following steps:

[0007] S1: Conduct behavioral analysis on electric vehicles based on order data to obtain the probability distribution of the starting charging time, the initial charging state of charge of electric vehicles, the probability distribution of charging power levels, and the charging time of electric vehicles under different environments;

[0008] S2: According to the behavior analysis results of step S1, the daily mileage, initial charging time and initial battery SOC of the electric vehicle are obtained by random simulation using the Monte Carlo method;

[0009] S3: According to the random simulation result of step S2 and based on the charging load calculation formula, the charging load at the physical level is predicted.

[0010] Preferably, in step S1, the functional expression of the probability distribution of the starting charging time is:

[0011]

[0012] Where: f h,s (t start ) is the starting charging time of the electric vehicle t start The probability distribution function of α, β and γ are the proportions of electric vehicles charging in the morning, middle and evening respectively; μ m,s , σ m,s , μ n,s , σ n,s , μ e,s and σ e,s are the starting charging time t of the electric vehicle start The expected value and standard deviation of the probability distribution of early, middle and late; t m ,t n ,t e are time variables with probability distributions of early, middle and late respectively.

[0013] Preferably, in step S1, the initial charging state of charge of the electric vehicle is calculated by the following formula:

[0014]

[0015] Where: SOC EV,0 The state of charge for the initial charging of electric vehicles; SOC en,EV is the state of charge of the electric vehicle before driving; d m E is the daily mileage of the electric vehicle when it reaches the charging station; EV is the power consumption of electric vehicles per 100 kilometers; C EV is the battery capacity of the electric vehicle; f en (SOC en,EV ) is SOC en,EV The probability distribution function of SOC en,EV The lower and upper limits of f m (d m ) is d m The probability distribution function of m , σ m They are d mThe expected value and standard deviation of a probability distribution.

[0016] Preferably, in step S1, the functional expression of the probability distribution of the charging power level is:

[0017]

[0018] Where: f EV (P rated ) is the charging power level P of the electric vehicle rated The probability distribution function of EV , σ EV They are respectively the charging power level P of electric vehicles rated The expected value and standard deviation of a probability distribution.

[0019] Preferably, in step S1, the charging time of the electric vehicle is calculated by the following formula:

[0020]

[0021] Where: t sp Charging time for electric vehicles; η EV Charging efficiency for electric vehicles; P rated Charging power levels for electric vehicles.

[0022] Preferably, in step S3, the charging load calculation formula is:

[0023]

[0024] Where: P EV,j (t) is the charging load; P rated,j is the rated charging power of the jth electric vehicle; t is the charging time; t start,j is the starting charging time of the jth electric vehicle; t sp,j is the charging time of the jth electric car.

[0025] As a preference, the method further comprises the following steps:

[0026] S4: Considering economic incentives, charging subsidies and user personality characteristics are integrated into the Logistic function to establish a user response model based on the Logistic function;

[0027] S5: Obtaining the correlation between the electric vehicle user response amount and the charging electricity price according to the user response model based on the Logistic function;

[0028] S6: Based on the association relationship and in combination with the charging load calculation formula, the charging load of the electric vehicle under economic incentives is predicted.

[0029] Preferably, in step S4, the user response model based on the Logistic function is:

[0030]

[0031] y t =a′x t +b′ (9)

[0032] Where: p t is the user response probability function; L is the proportional factor of the response probability; y t is the net input of the function; a′ is the price sensitivity factor; x t is the charging price; b′ is the random error.

[0033] Preferably, in step S6, the electric vehicle charging load under the economic incentive is calculated by the following formula:

[0034]

[0035] Where: is the charging load of electric vehicles under economic incentives; M is the number of electric vehicles randomly simulated by the Monte Carlo method; P EV,j (t) is the charging load.

[0036] The beneficial effects of the present invention are:

[0037] The present invention can predict and analyze the charging load of electric vehicles from physical perspectives such as region, season, weather, and electric vehicle type based on order data, and further analyze the relationship between electricity price and charging load from an economic perspective and predict the changes in electric vehicle charging load under economic stimulus, thereby improving the accuracy and generalization of load forecasting. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0039] Figure 1 It is a flow chart of a method for predicting charging load of electric vehicles based on order data of the present invention;

[0040] Figure 2 This is a schematic diagram of the electric vehicle load prediction process based on Monte Carlo simulation of the present invention;

[0041] Figure 3It is a schematic diagram of the overall framework of the electric vehicle charging load prediction method based on order data of the present invention;

[0042] Figure 4 This is a schematic diagram of the probability distribution result of the 24-hour starting charging time in a specific embodiment;

[0043] Figure 5 It is a schematic diagram of the probability distribution result of the power level of an electric vehicle according to a specific embodiment;

[0044] Figure 6 A schematic diagram of predicted charging loads of electric vehicles in different regions according to a specific embodiment;

[0045] Figure 7 It is a schematic diagram of the predicted charging load of electric vehicles in different seasons according to a specific embodiment;

[0046] Figure 8 It is a schematic diagram of predicted charging load of electric vehicles under different weather conditions according to a specific embodiment;

[0047] Fig. 9 It is a schematic diagram of electric vehicle charging prediction load under different electric vehicle types according to a specific embodiment;

[0048] Fig.10 A schematic diagram of the correlation between charging price and the number of user responses in a specific embodiment;

[0049] Fig.11 It is a schematic diagram of a curve showing the relationship between the actual charging price of a charging station and the number of user responses from 8:00 to 23:00 in a specific embodiment;

[0050] Fig.12 This is a schematic diagram of load changes at a charging station before and after charging prices are adjusted under peak load demand from 10 a.m. to 12 p.m. in a specific embodiment. DETAILED DESCRIPTION

[0051] The present invention is further described below in conjunction with the accompanying drawings and embodiments. It should be noted that, in the absence of conflict, the embodiments in this application and the technical features in the embodiments can be combined with each other. It should be noted that, unless otherwise specified, all technical and scientific terms used in this application have the same meanings as those generally understood by those of ordinary skill in the art to which this application belongs. The words "including" or "comprising" and the like used in the disclosure of the present invention mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0052] like Figure 1-3 As shown, the present invention provides a method for predicting charging load of electric vehicles based on order data, comprising the following steps:

[0053] S1: Conduct behavioral analysis on electric vehicles based on order data to obtain the probability distribution of the starting charging time, the initial charging state of charge of the electric vehicle, the probability distribution of the charging power level, and the charging time of the electric vehicle under different environments.

[0054] In a specific embodiment, the functional expression of the probability distribution of the starting charging time is:

[0055]

[0056] Where: f h,s (t start ) is the starting charging time of the electric vehicle t start The probability distribution function of α, β and γ are the proportions of electric vehicles charging in the morning, middle and evening respectively; μ m,s , σ m,s , μ n,s , σ n,s , μ e,s and σ e,s are the starting charging time t of the electric vehicle start The expected value and standard deviation of the probability distribution of early, middle and late; t m ,t n ,t e are time variables with probability distributions of early, middle and late respectively.

[0057] The starting charging time is random, and there is a certain probability of charging at the end of each trip of the user. Since a large number of independent random variables tend to be normally distributed after appropriate standardization, the probability distribution of the starting charging time can be fitted with a normal distribution. In the above embodiment, the present invention analyzes the actual travel data of electric vehicles, and the starting charging time of electric vehicles is mainly composed of three peaks in the morning, afternoon and evening, which is different from the starting charging time distribution of the prior art with only one peak. The starting charging time of the electric vehicle of the present invention can obtain a more accurate probability distribution of the starting charging time by representing it through three normal distributions.

[0058] In a specific embodiment, the initial charging state of charge of the electric vehicle is calculated by the following formula:

[0059]

[0060]

[0061] Where: SOC EV,0 The state of charge for the initial charging of electric vehicles; SOC en,EV is the state of charge of the electric vehicle before driving; d m E is the daily mileage of the electric vehicle when it reaches the charging station; EV is the power consumption of electric vehicles per 100 kilometers; CEV is the battery capacity of the electric vehicle; f en (SOC en,EV ) is SOC en,EV The probability distribution function of SOC en,EV The lower and upper limits of f m (d m ) is d m The probability distribution function of m , σ m They are d m The expected value and standard deviation of a probability distribution.

[0062] Daily mileage is one of the core factors in determining charging needs. Daily mileage is positively correlated with the power consumption of the electric vehicle and negatively correlated with the state of charge at the time of charging. In the above embodiment, the present invention establishes a daily mileage distribution probability model and uses the model as an important indicator for obtaining the state of charge at the time of charging, thereby improving the accuracy of the initial charging state of the electric vehicle of the present invention.

[0063] In a specific embodiment, the functional expression of the probability distribution of the charging power level is:

[0064]

[0065] Where: f EV (P rated ) is the charging power level P of the electric vehicle rated The probability distribution function of EV , σ EV They are respectively the charging power level P of electric vehicles rated The expected value and standard deviation of a probability distribution.

[0066] In the prior art, the charging power of electric vehicles is generally equivalent to the rated power of the charging pile, but the actual charging power depends on the minimum value of the rated power of the electric vehicle and the rated power of the charging pile. In the above embodiment, the present invention obtains the distribution of the charging power level of electric vehicles by fitting the actual charging station order data, which conforms to the normal distribution. Therefore, a more accurate probability distribution of the charging power level can be obtained through the function expression shown in formula (5).

[0067] In a specific embodiment, the charging time of the electric vehicle is calculated by the following formula:

[0068]

[0069] Where: t sp Charging time for electric vehicles; η EV Charging efficiency for electric vehicles; P ratedCharging power levels for electric vehicles.

[0070] S2: According to the behavior analysis results of step S1, the daily mileage, initial charging time and initial battery SOC of the electric vehicle are obtained by random simulation using the Monte Carlo method.

[0071] S3: According to the random simulation result of step S2 and based on the charging load calculation formula, the charging load at the physical level is predicted.

[0072] In a specific embodiment, the charging load calculation formula is:

[0073]

[0074] Where: P EV,j (t) is the charging load; P rated,j is the rated charging power of the jth electric vehicle; t is the charging time; t start,j is the starting charging time of the jth electric vehicle; t sp,j is the charging time of the jth electric car.

[0075] In a specific embodiment, the spatiotemporal distribution of electric vehicles in different regions, different times, and different weather conditions is extracted based on factors such as city area information, date information, and weather information. According to the spatiotemporal distribution of electric vehicles, the daily mileage, initial charging time, and initial battery SOC of each vehicle are randomly simulated. Then, the charging time of each vehicle is calculated based on formula (6), and the charging load P of the electric vehicle is calculated using the charging load calculation formula (7). EV,j (t). Finally, loop N (the number of vehicles in the target area), and after one day of simulation, add up the charging loads of all electric vehicles to output the total charging load of the area.

[0076] In a specific embodiment, the electric vehicle charging load prediction method based on order data of the present invention further includes the following steps:

[0077] S4: Considering economic incentives, charging subsidies and user personality characteristics are integrated into the Logistic function, and a user response model based on the Logistic function is established.

[0078] In a specific embodiment, the user response model based on the Logistic function is:

[0079]

[0080] y t =a′x t +b′ (9)

[0081] Where: pt is the user response probability function; L is the proportional factor of the response probability; y t is the net input of the function; a′ is the price sensitivity factor; x t is the charging price; b′ is the random error.

[0082] The charging decision of electric vehicle users is also affected by factors such as charging subsidies and user personality traits. The higher the charging subsidy or the lower the charging price a user receives at a certain moment, the higher the probability of charging at the corresponding moment. Conversely, the lower the charging subsidy or the higher the charging price (the subsidy price is negative), the lower the probability of the user charging at the corresponding moment. The Logistic function is a monotonically increasing function with a value interval of (0,1). When the value approaches negative infinity, the value is 0, and when the value approaches positive infinity, the value is 1. Therefore, the present invention uses it as the basis of the user response model. When in use, the charging subsidy and user personality traits are integrated into the Logistic function, and considering the different sensitivity of different users to prices, a price sensitivity factor is introduced to establish a user response model based on the Logistic function as shown in formulas (8)-(9).

[0083] S5: Obtaining the correlation between the electric vehicle user response amount and the charging electricity price according to the user response model based on the Logistic function.

[0084] S6: Based on the association relationship and in combination with the charging load calculation formula, the charging load of the electric vehicle under economic incentives is predicted.

[0085] In a specific embodiment, the electric vehicle charging load under the economic incentive is calculated by the following formula:

[0086]

[0087] Where: is the charging load of electric vehicles under economic incentives; M is the number of electric vehicles randomly simulated by the Monte Carlo method; P EV,j (t) is the charging load.

[0088] In a specific embodiment, Zhengzhou City, Henan Province is taken as a research area, and based on the actual travel data of the area, the electric vehicle charging load prediction method based on order data of the present invention is used to predict the electric vehicle charging load.

[0089] In this embodiment, the probability distribution of the starting charging time under different environments is as follows: Figure 4 As shown in Figure 2, the probability distribution of charging power levels under different environments is as follows: Figure 5 As shown in Figure 2, the predicted charging load of electric vehicles in different regions, seasons, weather conditions, and electric vehicle types is as follows: Figure 6-9As shown in Table 1, the logistic relationship between the charging price and the number of user responses from 8:00 to 23:00 is shown in Table 1. Fig.10 As shown in the figure, the relationship curve between the actual charging price of the charging station and the number of user responses from 8:00 to 23:00 is as follows: Fig.11 As shown in Figure 2, the load changes of the charging station before and after the charging price adjustment under the peak load demand from 10:00 to 12:00 are as follows: Fig.12 shown.

[0090] Table 1 Parameters of the logistics relationship between charging price and user response quantity from 8:00 to 23:00

[0091]

[0092] In summary, the present invention predicts and analyzes the charging load of electric vehicles from the physical perspectives of region, season, weather, electric vehicle type, etc. based on order data, and predicts the relationship between time-of-use electricity price and charging load from the economic level and predicts the change of electric vehicle charging load under economic stimulus, which can provide a basis for the formulation of electric vehicle-related policies, time-of-use electricity price and service fee. Compared with the prior art, the present invention has significant progress.

[0093] The above description is only a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any technician familiar with the profession can make some changes or modifications to equivalent embodiments of equivalent changes using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of the technical solution of the present invention.

Claims

1. A method for predicting electric vehicle charging load based on order data, characterized in that: The following steps are involved: S1: Conduct behavioral analysis on electric vehicles based on order data to obtain the probability distribution of the starting charging time, the initial charging state of charge of electric vehicles, the probability distribution of charging power levels, and the charging time of electric vehicles under different environments; S2: According to the behavior analysis results of step S1, the daily mileage, initial charging time and initial battery SOC of the electric vehicle are obtained by random simulation using the Monte Carlo method; S3: According to the random simulation result of step S2 and based on the charging load calculation formula, the charging load at the physical level is predicted.

2. The electric vehicle charging load prediction method based on order data according to claim 1 is characterized in that: In step S1, the function expression of the probability distribution of the starting charging time is: Where: f h,s (t start ) is the starting charging time of the electric vehicle t start The probability distribution function of α, β and γ are the proportions of electric vehicles charging in the morning, middle and evening respectively; μ m,s , σ m,s , μ n,s , σ n,s , μ e,s and σ e,s are the starting charging time t of the electric vehicle start The expected value and standard deviation of the probability distribution of early, middle and late; t m ,t n ,t e are time variables with probability distributions of early, middle and late respectively.

3. The electric vehicle charging load prediction method based on order data according to claim 1 is characterized in that: In step S1, the initial charging state of charge of the electric vehicle is calculated by the following formula: Where: SOC EV,0 The state of charge for the initial charging of electric vehicles; SOC en,EV is the state of charge of the electric vehicle before driving; d m E is the daily mileage of the electric vehicle when it reaches the charging station; EV C is the power consumption of electric vehicles per 100 kilometers; EV is the battery capacity of the electric vehicle; f en (SOC en,EV ) is SOC en,EV The probability distribution function of SOC en,EV The lower and upper limits of f m (d m ) is d m The probability distribution function of m , σ m They are d m The expected value and standard deviation of a probability distribution.

4. The electric vehicle charging load prediction method based on order data according to claim 1 is characterized in that: In step S1, the functional expression of the probability distribution of the charging power level is: Where: f EV (P rated ) is the charging power level P of the electric vehicle rated The probability distribution function of EV , σ EV They are respectively the charging power level P of electric vehicles rated The expected value and standard deviation of a probability distribution.

5. The electric vehicle charging load prediction method based on order data according to claim 1 is characterized in that: In step S1, the charging time of the electric vehicle is calculated by the following formula: Where: t sp Charging time for electric vehicles; η EV Charging efficiency for electric vehicles; P rated Charging power levels for electric vehicles.

6. The electric vehicle charging load prediction method based on order data according to claim 1 is characterized in that: In step S3, the charging load calculation formula is: Where: P EV,j (t) is the charging load; P rated,j is the rated charging power of the jth electric vehicle; t is the charging time; t start,j is the starting charging time of the jth electric vehicle; t sp,j is the charging time of the jth electric car.

7. The electric vehicle charging load prediction method based on order data according to any one of claims 1 to 6, characterized in that: The following steps are also included: S4: Considering economic incentives, charging subsidies and user personality characteristics are integrated into the Logistic function to establish a user response model based on the Logistic function; S5: Obtaining the correlation between the electric vehicle user response amount and the charging electricity price according to the user response model based on the Logistic function; S6: Based on the association relationship and in combination with the charging load calculation formula, the charging load of the electric vehicle under economic incentives is predicted.

8. The electric vehicle charging load prediction method based on order data according to claim 7 is characterized in that: In step S4, the user response model based on the Logistic function is: y t =a′x t +b′ (9) Where: p s is the user response probability function; L is the proportional factor of the response probability; y t is the net input of the function; a′ is the price sensitivity factor; x t is the charging price; b′ is the random error.

9. The electric vehicle charging load prediction method based on order data according to claim 8 is characterized in that: In step S6, the electric vehicle charging load under the economic incentive is calculated by the following formula: Where: is the charging load of electric vehicles under economic incentives; M is the number of electric vehicles randomly simulated by the Monte Carlo method; P EV,j (t) is the charging load.

Images