Intelligent guidance method and device for commuting charging stop driven by user aging preference

Abstract

The invention discloses a user aging preference driven commuting charging stopping intelligent guiding method and device. The method comprises the steps of constructing a charging stop data set by using collected behavior data, training a new energy user aging preference portrait model according to the charging stop data set, and then collecting behavior data before the current moment and inputting the behavior data into the new energy user aging preference portrait model for prediction. And according to the prediction result and the collected commuting parking area planning data, the comprehensive commuting parking charging convenience degree of all the users is obtained, an intelligent guiding parking space model is constructed according to the comprehensive commuting parking charging convenience degree, and a genetic algorithm is adopted for solving to obtain the parking space position corresponding to the new energy automobile of each user. According to the invention, the advantage of carrying out commuting charging stop scheduling on all new energy vehicles comprehensively is realized, the effect of considering the comprehensive commuting charging stop convenience and timeliness of all users is achieved, and the target of optimal commuting charging stop comprehensive arrangement is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of commuting charging scheduling for new energy vehicles, and particularly relates to a commuting charging intelligent guidance method and device driven by user time preference. Background Art

[0002] In recent years, the new energy vehicle industry has developed rapidly under the impetus of technology and policies and has become an important part of urban transportation. Technological iteration has improved industry standards, and measures such as purchase subsidies have promoted the rapid growth of the vehicle ownership.

[0003] The global green travel revolution has promoted the transformation of traditional parking lots, and charging piles have been rapidly popularized in commercial areas, residential areas, highway service areas, etc. With the increasing demand for charging spaces, how to reasonably arrange parking and charging and intelligently guide users' commuting has become an urgent problem.

[0004] Most of the existing parking lots do not have a commuting charging scheduling strategy for new energy vehicle owners, which easily causes considerable inconvenience to some new energy vehicle owners during the commuting charging process, affecting the efficiency of users' commuting charging and the satisfaction of users with the parking lot. For example, although some parking lots are equipped with charging piles, due to the lack of reasonable guidance and management, the charging piles are often occupied by fuel vehicles, or the utilization rate of the charging piles is low, resulting in new energy vehicle owners being unable to find available charging facilities in time. This situation not only affects the travel experience of vehicle owners but also exacerbates the waste of charging resources. Summary of the Invention

[0005] To solve the problems in the background art, the present invention provides a commuting charging intelligent guidance method and device driven by user time preference, which solves the problem of the lack of intelligent scheduling for new energy vehicle owners in the prior art.

[0006] The technical solutions adopted by the present invention include: 1. A commuting charging intelligent guidance method driven by user time preference: S1. Collect the behavior data of each new energy vehicle entering and leaving the parking lot in a number of historical days, and construct a charging and parking data set according to the behavior data of all new energy vehicles.

[0007] S2. Construct a new energy vehicle user time preference portrait model in a computer, input the charging and parking data set into the new energy vehicle user time preference portrait model for training, and obtain a trained new energy vehicle user time preference portrait model.

[0008] S3. Input the collected behavior data before the current moment into the new energy vehicle user time-efficiency preference portrait model for prediction. According to the prediction results and the collected commuting parking area planning data within the preset sliding window time, the total commuting stop-and-charge convenience of each new energy vehicle is obtained. According to the total commuting stop-and-charge convenience, the comprehensive commuting stop-and-charge convenience of all users is obtained.

[0009] S4. Build an intelligent guidance parking space model based on the comprehensive commuting, stopping and charging convenience, use a genetic algorithm to process the intelligent guidance parking space model to obtain the parking space position corresponding to each user's new energy vehicle, and guide each new energy vehicle to enter the corresponding parking space according to the obtained parking space position.

[0010] The S1 is specifically: S11. Collect the behavior data of each new energy vehicle entering and exiting the parking lot every time within a few days.

[0011] S12. Count the probability values ​​of each new energy vehicle leaving the parking lot from each parking lot exit on all days based on the behavior data, and use the probability values ​​of leaving the parking lot from all parking lot exits as probability data.

[0012] S14. Use the behavior data as input data and the corresponding probability data as labels to construct the charging and stopping data of each new energy vehicle.

[0013] S15. The charging stop data of all new energy vehicles are aggregated to obtain a charging stop data set.

[0014] The behavior data of each new energy vehicle entering and exiting the parking lot each time in S11 includes the user number of the new energy vehicle, the parking lot entrance number when entering the parking lot, the time period when entering the parking lot entrance, and the parking lot exit number when leaving the parking lot.

[0015] The new energy vehicle user time preference portrait model in S2 adopts a long short-term memory neural network.

[0016] The S3 is specifically: S31. Collect the behavior data of all new energy vehicles in a preset time period before the current moment, and input the behavior data into the trained new energy vehicle user time preference portrait model to predict the probability value of each new energy vehicle leaving the parking lot from each parking lot exit within the preset sliding window time after the current moment.

[0017] S32. Collect the commuting parking area planning data of each new energy vehicle within a preset sliding window time after the current moment, and obtain the total commuting parking and charging convenience of each new energy vehicle based on the commuting parking area planning data and the probability value of each new energy vehicle leaving the parking lot from each parking lot exit.

[0018] S33. Obtain the comprehensive commuting and charging convenience of all users based on the total commuting and charging convenience of each new energy vehicle user.

[0019] The commuting parking area planning data of each new energy vehicle in S32 includes the user number of the new energy vehicle, the time period when entering the parking lot entrance, the parking lot entrance number when entering the parking lot, the number of the initially estimated parking space, the difficulty value of driving from the parking lot entrance to the parking space, the difficulty value of driving from the parking space to the parking lot exit, the total number of parking lot exits, the power of the new energy vehicle when entering the parking lot entrance, and the density of new energy vehicles in the parking space.

[0020] The total commuting and charging convenience of each new energy vehicle in S32 is obtained by the following formula: ρ i = 0.5α i + 0.3β i + 0.2γ i ; α i = 1 - (D i，INj (P)+ ∑ k=1 n f i，INj，OUTk × D i，OUTk (P)) / 2; β i =(1 - D i，INj (P)) × (1 - SOC / 100); γ i = 1 - INT(P).

[0021] Among them, i represents the user number of the new energy vehicle; ρ i represents the total commuting and charging convenience of the new energy vehicle with user number i; α i represents the commuting and charging convenience of the new energy vehicle with user number i based on the parking and pick-up of new energy vehicle users; β i represents the commuting and charging convenience of the new energy vehicle with user number i based on the charging of new energy vehicles; γ i represents the commuting and charging convenience of the new energy vehicle with user number i based on the density of new energy vehicles; j represents the index and also the parking lot entrance number when entering the parking lot; D i，INj (P) represents the difficulty value of the new energy vehicle with user number i driving from the parking lot entrance numbered j to the parking space numbered P; k represents the index and also the parking lot exit number when leaving the parking lot; D i，OUTk (P) represents the difficulty value of the new energy vehicle with user number i driving from the parking space numbered P to the parking lot exit numbered k; n represents the total number of parking lot exits;f i，INj，OUTk It represents the probability value that a new energy vehicle with user number i enters the entrance of a parking lot with number j and then leaves the parking lot from the exit of the parking lot with number k; SOC represents the power of the new energy vehicle when it enters the entrance of the parking lot; INT(P) represents the density of new energy vehicles in the parking space with number P.

[0022] Specifically, S4 is as follows: S41. Binary code the numbers of all parking spaces to obtain the numbers of the parking spaces after binary coding.

[0023] S42. Determine the constraint conditions according to the numbers of the parking spaces after binary coding, and construct an intelligent guiding parking space model with the comprehensive commuting charging convenience as the fitness function.

[0024] S43. Use the genetic algorithm to process the intelligent guiding parking space model, obtain the combination of the parking space positions of all users' new energy vehicles under the maximum comprehensive commuting charging convenience, and obtain the corresponding parking space positions of each user's new energy vehicle according to the combination of the parking space positions.

[0025] S44. Guide each new energy vehicle to enter the corresponding parking space position according to the parking space positions obtained in S43.

[0026] The intelligent guiding parking space model in S42 is set according to the following formula: Max(ρ all (P all )); ρ all (P all )=∑ i=1 m ρ i ; 0<P ibin <P max ; P ibin ≠P jbin 。

[0027] Among them, i represents the user number of the new energy vehicle; ρ i represents the total commuting charging convenience of the new energy vehicle with user number i; m represents the total number of users of new energy vehicles that need commuting charging intelligent guidance; P all represents the combination of the parking space positions of all users' new energy vehicles; ρ all (P all ) represents the comprehensive commuting charging convenience of all users' new energy vehicles under the combination of P all ; Max(ρ all (P all)) represents maximizing the fitness function; P ibin represents the binary encoding of the parking space corresponding to the new energy vehicle with user number i; P jbin represents the number after binary encoding of the parking space corresponding to the new energy vehicle with user number j; P max represents the maximum value among the numbers after binary encoding of all parking spaces.

[0028] II. A commuting charging intelligent guidance device driven by user time - efficiency preference: It includes: an information acquisition module for collecting the behavior data of new energy vehicles and the commuting parking area planning data; an information storage module for storing the behavior data of new energy vehicles and the commuting parking area planning data; a model library module for constructing and storing the new energy vehicle user time - efficiency preference portrait model and the intelligent guidance parking space model; a calculation module for processing and calculating the total commuting charging convenience of each new energy vehicle, the comprehensive commuting charging convenience of all users, and the combination of the parking space positions of new energy vehicles of all users under the maximized comprehensive commuting charging convenience; a visual guidance module for guiding each new energy vehicle to drive to the corresponding parking space position.

[0029] The innovation of the present invention lies in adopting the new energy vehicle user time - efficiency preference portrait model and the intelligent guidance parking space model, realizing the advantage of comprehensively scheduling commuting charging for all new energy vehicles, achieving the effect of considering the comprehensive commuting charging convenience and timeliness of all users, and reaching the goal of the optimal comprehensive arrangement of commuting charging.

[0030] The beneficial effects of the present invention are: 1. Based on commuting charging, the present invention introduces the new energy vehicle user time - efficiency preference portrait model and the intelligent guidance parking space model. By calculating the total commuting charging convenience of new energy vehicles of each user and the comprehensive commuting charging convenience of all users, it realizes the comprehensive scheduling of commuting charging for all new energy vehicles.

[0031] 2. When meeting the commuting charging requirements of new energy vehicle owners, the present invention takes into account the comprehensive commuting charging convenience and timeliness of all users, achieving the optimal comprehensive arrangement of commuting charging. Description of the Drawings

[0032] Figure 1 is the flowchart of the method of the present invention.

[0033] Figure 2 is the exemplary layout diagram of the first parking lot of the present invention.

[0034] Figure 3 is the exemplary layout diagram of the second parking lot of the present invention.

[0035] Figure 4 Schematic diagram of the main functional modules of the device of the present invention.

[0036] Figure 5 Flowchart of the genetic algorithm used in the method of the present invention. Specific embodiments

[0037] The present invention will be described in more detail below with reference to the accompanying drawings and embodiments. However, the present invention is not limited thereto. For those of ordinary skill in the art in the technical field, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements are also considered within the protection scope of the present invention. The content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

[0038] Specific embodiments of the present invention are as follows: Embodiment 1:

[0039] As Figure 1 shown, the commuting parking and charging intelligent guidance method of this embodiment includes the following steps: S1. Collect the behavior data of each new energy vehicle entering and leaving the parking lot several times within a number of historical days, and construct a parking and charging data set according to the behavior data of all new energy vehicles.

[0040] S11. Collect the behavior data of each new energy vehicle entering and leaving the parking lot each time within a number of historical days.

[0041] The behavior data of each new energy vehicle entering and leaving the parking lot each time includes the user number of the new energy vehicle, the parking lot entrance number when entering the parking lot, the time period when entering the parking lot entrance, and the parking lot exit number when leaving the parking lot.

[0042] Specifically, form a data group with the user number of the new energy vehicle, the parking lot entrance number when entering the parking lot, the time period when entering the parking lot, and the parking lot exit number when leaving the parking lot as the behavior data.

[0043] S12. According to the behavior data, statistically calculate the probability value of each new energy vehicle leaving the parking lot from each parking lot exit within all days, and combine the probability values of leaving the parking lot from all parking lot exits as probability data.

[0044] In specific implementation, the probability value of leaving the parking lot from each parking lot exit is the number of times the new energy vehicle leaves the parking lot from the corresponding parking lot exit divided by the total number of times leaving the parking lot from all parking lot exits. The sum of the probability values of all parking lot exits for each new energy vehicle leaving the parking lot is one.

[0045] S14. Use the behavior data as input data and the corresponding probability data as labels to construct the parking and charging data of each new energy vehicle.

[0046] S15. The charging and parking data of all new energy vehicles are summarized to obtain a charging and parking data set.

[0047] S2. Build a new energy vehicle user time - effect preference portrait model in a computer, input the charging and parking data set into the new energy vehicle user time - effect preference portrait model for training, and obtain a trained new energy vehicle user time - effect preference portrait model.

[0048] The new energy vehicle user time - effect preference portrait model adopts a long - short - term memory neural network.

[0049] Specifically, the behavior data is used as the input data of the new energy vehicle user time - effect preference portrait model, and the corresponding probability data is used as the label data of the new energy vehicle user time - effect preference portrait model and input into the new energy vehicle user time - effect preference portrait model for training.

[0050] S3. Input the behavior data collected before the current moment into the new energy vehicle user time - effect preference portrait model for prediction. According to the prediction results and the commuting parking area planning data within the preset sliding window time after the current moment, obtain the total commuting charging and parking convenience of each new energy vehicle, and obtain the comprehensive commuting charging and parking convenience of all users based on the total commuting charging and parking convenience.

[0051] S31. Collect the behavior data of all new energy vehicles within the preset time period before the current moment, and input the behavior data into the trained new energy vehicle user time - effect preference portrait model to predict the probability value of each new energy vehicle leaving the parking lot from each parking lot exit within the preset sliding time after the current moment.

[0052] S32. Collect the commuting parking area planning data of each new energy vehicle within the preset sliding window time after the current moment, and obtain the total commuting charging and parking convenience of each new energy vehicle according to the commuting parking area planning data and the probability value of each new energy vehicle leaving the parking lot from each parking lot exit obtained in S31.

[0053] Furthermore, the preset time period after the current moment predicted is longer than the preset sliding window time after the current moment.

[0054] The commuting parking area planning data of each new energy vehicle includes the user number of the new energy vehicle, the time period when entering the parking lot entrance, the parking lot entrance number when entering the parking lot, the preliminary estimated parking space number, the difficulty value of driving from each parking lot entrance to all parking spaces, the difficulty value of driving from each parking space to all parking lot exits, the total number of parking lot exits, the battery power of the new energy vehicle when entering the parking lot entrance, and the new energy vehicle density of each parking space.

[0055] Specifically, the difficulty values for driving from each parking lot entrance to all parking spaces and from each parking space to all parking lot exits are constants, which are obtained based on experience after the area design of the parking lot. When entering the parking lot entrance, the power of the new energy vehicle is obtained from the instrument panel of the new energy vehicle through the data link. The density of new energy vehicles in each parking space is the proportion of new energy vehicles parked in the Y parking spaces around the parking space. For example, if there are 5 new energy vehicles parked in the 10 parking spaces around a parking space, the proportion is 0.5, that is, the density of new energy vehicles in this parking space is 0.5. The density of new energy vehicles in each parking space is obtained by scanning with the monitoring camera on the parking space.

[0056] The total commuting charging convenience of each new energy vehicle is processed according to the following formula: ρ Ui，INj，T，P =0.5α Ui，INj，T，P +0.3β Ui，INj，T，P +0.2γ Ui，INj，T，P ; α Ui，INj，T，P =1-(D INj (P)+∑ k=1 n f Ui，INj，OUTk，T × D OUTk (P)) / 2; β Ui，INj，T，P =(1-D INj (P))×(1-SOC Ui / 100); γ Ui，INj，T，P =1-INT(P).

[0057] Where, i represents the user number of the new energy vehicle; j represents the parking lot entrance number when entering the parking lot; Ui represents the new energy vehicle with user number i; INj represents the parking lot entrance with number j; T represents the time period when entering the parking lot entrance; P represents the preliminary estimated parking space number; ρ Ui，INj，T，P represents the total commuting charging convenience of the new energy vehicle with user number i entering the parking lot entrance with number j and parking in the parking space with number P during the T time period; α Ui，INj，T，P represents the commuting charging convenience based on the parking and pick-up of new energy vehicle users of the new energy vehicle with user number i entering the parking lot entrance with number j and parking in the parking space with number P during the T time period; β Ui，INj，T，P represents the commuting charging convenience based on the charging of new energy vehicles of the new energy vehicle with user number i entering the parking lot entrance with number j and parking in the parking space with number P during the T time period; γ Ui，INj，T，PIt represents the commuting charging convenience based on the new energy vehicle density when a new energy vehicle with user number i enters the entrance of parking lot numbered j and parks at parking space numbered P during time period T; D INj (P) represents the difficulty value of driving from the entrance of parking lot numbered j to parking space numbered P; n represents the total number of parking lot exits; k represents the index and also represents the parking lot exit number when leaving the parking lot; f Ui，INj，OUTk，T It represents the probability value when a new energy vehicle with user number i leaves the parking lot from the parking lot exit numbered k on the premise that the new energy vehicle with user number i enters the entrance of parking lot numbered j during time period T; D OUTk (P) represents the difficulty value of driving from the parking space at position P to the parking lot exit numbered k; SOC Ui It represents the battery level of the new energy vehicle when the new energy vehicle with user number i enters the parking lot entrance, taking the value in percentage; INT(P) represents the new energy vehicle density of the parking space numbered P.

[0058] In specific implementation, the commuting charging intelligent guidance of this method occurs after the new energy vehicle enters the parking lot entrance, so the entrance number of the new energy vehicle for each user is determined.

[0059] S33. Within the preset sliding window time after the current moment, obtain the comprehensive commuting charging convenience of all users according to the total commuting charging convenience of each new energy vehicle user.

[0060] The comprehensive commuting charging convenience of all users is set according to the following formula: ρ all (P all ) = ∑ i=1 m ρ Ui，INj，T，P .

[0061] Among them, i represents the user number of the new energy vehicle; j represents the parking lot entrance number when entering the parking lot; T represents the time period when entering the parking lot entrance; P represents the parking space number; U i represents the new energy vehicle with user number i; INj represents the parking lot entrance numbered j; ρ Ui，INj，T，P represents the total commuting charging convenience when a new energy vehicle with user number i enters the entrance of parking lot numbered j and parks at parking space numbered P during time period T; m represents the total number of new energy vehicles of users who need commuting charging intelligent guidance; P all represents the combination of the parking space positions of the new energy vehicles of all users; ρ all (P all ) represents the comprehensive commuting charging convenience of the new energy vehicles of all users in the combination of P all .

[0062] S4. Construct an intelligent guiding parking space model based on the comprehensive commuting charging convenience, and use the genetic algorithm to process the intelligent guiding parking space model to obtain the parking space positions corresponding to the new energy vehicles of each user, and guide each new energy vehicle into the corresponding parking space according to the obtained parking space positions.

[0063] S41. Binary encode the numbers of all parking spaces to obtain the numbers of the parking spaces after binary encoding.

[0064] S42. Determine the constraint conditions according to the numbers of the parking spaces after binary encoding, and construct an intelligent guiding parking space model with the comprehensive commuting charging convenience as the fitness function.

[0065] The intelligent guiding parking space model is set according to the following formula: Max(ρ all (P all )); ρ all (P all ) = ∑ i=1 m ρ Ui，INj，T，P ; 0 < P ibin < P max ; P ibin ≠ P jbin .

[0066] Among them, i represents the user number of the new energy vehicle; j represents the parking entrance number when entering the parking lot; T represents the time period when entering the parking lot entrance; P represents the number of the parking space; U i represents the new energy vehicle of the user with the user number i; INj represents the parking lot entrance with the number j; ρ Ui，INj，T，P represents the total commuting charging convenience of the new energy vehicle of the user with the user number i entering the parking lot entrance with the number j and parking in the parking space with the number P during the time period T; m represents the total number of new energy vehicles that need intelligent guiding for commuting charging; P all represents the combination of the parking space positions of the new energy vehicles of all users within the preset sliding window time; ρ all (P all ) represents the fitness function, that is, the comprehensive commuting charging convenience of the new energy vehicles of all users under the combination of P all ; Max(ρ all (P all )) represents maximizing the fitness function, that is, maximizing the comprehensive commuting charging convenience of the new energy vehicles of all users under the combination of P all ; P ibinThe binary code corresponding to the parking space for the new energy vehicle with user number i; P jbin The number after binary encoding of the parking space corresponding to the new energy vehicle with user number j; P max Represents the maximum value among the numbers after binary encoding of all parking spaces.

[0067] In specific implementation, P all Is a combination formed by selecting a parking space for each user for commuting charging and parking. Each user selects a different parking space for commuting charging and parking, constituting multiple P all .

[0068] S43. Use the genetic algorithm to process the intelligent guidance parking space model to obtain the combination of the parking space positions of all users' new energy vehicles under the maximized comprehensive commuting charging and parking convenience, and obtain the parking space positions corresponding to each user's new energy vehicle according to the combination of the parking space positions.

[0069] In specific implementation, as Figure 5 shown, the genetic algorithm is specifically as follows: Initialize the population: Generate a set of possible parking combinations.

[0070] Fitness evaluation: Calculate the comprehensive commuting charging and parking convenience of each parking space combination.

[0071] Selection: Select better parking space combinations according to the fitness values.

[0072] Crossover: Perform crossover operations on the selected parking space combinations to generate new combinations.

[0073] Mutation: Perform mutation operations on some parking space combinations to introduce randomness.

[0074] Iteration: Repeat the above steps until the optimal solution is found or the maximum number of iterations is reached.

[0075] S44. Guide each new energy vehicle into the corresponding parking space position according to the parking space positions obtained in S43.

[0076] Furthermore, repeat S3 - S4 to achieve continuous intelligent guidance for the commuting charging and parking of new energy vehicles.

[0077] As Figure 4As shown in the figure, the device of the present invention includes an information collection module for collecting the behavior data of new energy vehicles and the commuting parking area planning data; an information storage module for storing the behavior data of new energy vehicles and the commuting parking area planning data; a model library module for constructing and storing the time-effective preference portrait model of new energy vehicle users and the intelligent guiding parking space model; a calculation module for processing and calculating the total commuting charging convenience of each new energy vehicle, the comprehensive commuting charging convenience of all users, and the combination of the parking space positions of new energy vehicles of all users under the maximized comprehensive commuting charging convenience; and a visual guiding module for guiding each new energy vehicle to drive to the corresponding parking space position.

[0078] Embodiment 2:

[0079] This embodiment is implemented by the same method as Embodiment 1. During the implementation process: In this embodiment, the behavior data of a total of 200 new energy vehicles within one month of history is collected. As Figure 2 and Figure 3 shown, they are two exemplary layout diagrams of parking lots; this embodiment is implemented using the Figure 2 parking lot. The parking lot has a total of three entrances and exits. Taking the new energy vehicle with user number 1 as an example, the input data of the time-effective preference portrait model of new energy vehicle users is [U1, IN1, 13:00, OUT1], [U1, IN1, 13:00, OUT2], [U1, IN1, 13:00, OUT3], and the output is f U1，IN1，OUT1，13:00 = 0.4, f U1，IN1，OUT2，13:00 = 0.2, f U1，IN1，OUT1，13:00 = 0.4. Therefore, the output of the time-effective preference portrait model of new energy vehicle users is [0.4, 0.2, 0.4]. In this embodiment, the time period of 13:00 represents the time period from 13:00 to 14:00.

[0080] In addition to the new energy vehicle with user number 1, the new energy vehicle with user number 2 and the new energy vehicle with user number 3 also enter the parking lot during the time period from 13:00 to 14:00. Similarly, the output of the time-effective preference portrait model of the new energy vehicle with user number 2 is [0.6, 0.2, 0.2], and the output of the time-effective preference portrait model of the new energy vehicle with user number 3 is [0.2, 0.3, 0.5].

[0081] The new energy vehicle with user number 1 enters the parking lot through the entrance numbered 1 of the parking lot during the time period from 13:00 to 14:00, and the preliminary estimated parking position is P 23 , and the following formula is used for processing to obtain the total commuting charging convenience of the new energy vehicle with user number 1: ρ U1，IN1，13:00，P37 = 0.5α U1，IN1，13:00，P37 + 0.3β U1，IN1，13:00，P37 + 0.2γ U1，IN1，13:00，P37 ; α U1，IN1，13:00，P37 = 1 - D IN1 (P 37 ) + ∑ k=1 3 f U1，IN1，OUTk，13:00 × D OUTk (P 37 ) / 2; β U1，IN1，13:00，P37 =(1 - D IN1 (P 37 )) × (1 - SOC U1 / 100); γ U1，IN1，13:00，P37 = 1 - INT(P 37 );

[0082] Among them, the difficulty value D 37 (P IN1 ) from the entrance of Parking Lot No. 1 to parking space P 37 is 0.3, the difficulty value D 37 (P OUT1 ) from parking space P 37 to the exit of Parking Lot No. 1 is 0.5, the difficulty value D 37 (P OUT2 ) from parking space P 37 to the exit of Parking Lot No. 2 is 0.4, the difficulty value D 37 (P OUT3 ) from parking space P 37 to the exit of Parking Lot No. 3 is 0.7, and f U1，IN1，OUTk，13:00 is [0.4, 0.2, 0.4]. The power SOC U1 of the new energy vehicle when entering the entrance of Parking Lot No. 1 is 60, and the density of new energy vehicles in parking space P 37 is 0.2.

[0083] Therefore, α U1，IN1，13:00，P37 = 0.57, β U1，IN1，13:00，P37 = 0.28, γ U1，IN1，13:00，P37 = 0.8. Furthermore, the total commuting, parking, and charging convenience ρ 37 for the new energy vehicle with user ID 1 from the entrance of Parking Lot No. 1 to parking space P U1，IN1，13:00，P37 at the 13:00 - 14:00 time period is 0.529.

[0084] Since the intelligent guiding for commuting and charging stops after the new energy vehicle enters the parking lot entrance, the entrance number of the new energy vehicle for each user is determined.

[0085] Therefore, by the same token, it is obtained that during the time period from 13:00 to 14:00, the new energy vehicle with user number 2 is expected to drive from the parking lot entrance numbered 2 to P 15 The total commuting and charging convenience degree of the parking space is ρ U2，IN2，13:00，P15 = 0.433; during the time period from 13:00 to 14:00, the new energy vehicle with user number 3 drives from the parking lot entrance numbered 1 to P 50 The total commuting and charging convenience degree of the parking space is ρ U3，IN1，13:00，P50 = 0.561.

[0086] An intelligent guiding parking space model is constructed according to the comprehensive commuting and charging convenience degree. The constructed intelligent guiding parking space model is set according to the following formula: Max(ρ all (P all )); ρ all (P all ) = ∑ i=1 3 ρ Ui，INj，13:00，P ; 0 < P ibin < P max ; P ibin ≠ P jbin .

[0087] Among them, ρ Ui，INj，13:00，P are respectively: ρ U1，IN1，13:00，P37 , ρ U2，IN2，13:00，P15 and ρ U3，IN1，13:00，P50 .

[0088] First, code the parking spaces and use the genetic algorithm to process the intelligent guiding parking space model to obtain the parking space positions corresponding to the new energy vehicles of each user. The specific steps of coding and genetic algorithm processing are as follows: 1. Coding: Each parking space position is represented by a 6-bit binary code (because there are 63 parking spaces in the parking lot, 2^6 = 64 is sufficient to cover).

[0089] For the new energy vehicles of 3 users, the binary code length of the parking space combination P all is 12 bits (6 bits × 2). For example, P all = [P 37 , P 15 , P 50 has a binary code of 100101001111110010, where: P 37The binary code of is 100101, P 15 The binary code of is 001111 and P 50 The binary code of is 110010.

[0090] 2. Constraints: The binary code of each parking space location must satisfy: 0 < P ibin < P max ; P ibin ≠ P jbin .

[0091] Among them, P ibin represents the binary code of the parking space corresponding to the new energy vehicle with user number i; P jbin represents the binary code of the parking space corresponding to the new energy vehicle with user number j; P max represents the maximum value of the binary code of the parking space, that is, 1111111.

[0092] 3. Fitness function: The fitness function is the comprehensive commuting charging convenience ρ all (P all ), and the calculation formula is: ρ all (P all ) = ∑ i=1 3 ρ Ui，INj，13:00，P .

[0093] The goal is to optimize the parking space combination P all , so that ρ all (P all ) is maximized.

[0094] 4. Genetic algorithm process: 4.1. Initialize the population: Randomly generate a group of binary codes of parking space combinations, for example: Combination 1: 011110001100001110 (P 30 , P 12 , P 14 ), Combination 2: 010101100011001101 (P 21 , P 35 , P 13 ), Combination 3: 101010010101010001 (P 42 , P 21 , P 17 ).

[0095] 4.2. Fitness evaluation: Calculate the comprehensive commuting charging convenience ρ of each parking space combination all(P all )。

[0096] 4.3. Selection: Select better parking space combinations according to the fitness value. For example, select combination 1 and combination 2.

[0097] 4.4. Crossover: Perform crossover operations on the selected parking space combinations. For example: combination 1: 011110|001100|001110, combination 2: 010101|100011|001101. After crossover, new combinations are generated: 011110|100011|001110 and 010101|001100|001101.

[0098] 4.5. Mutation: Perform random mutations on the newly generated parking space combinations. For example, change the 3rd bit of 011110|100011|001110 from 1 to 0 to get 010110|100011|001110.

[0099] 4.6. Iteration: Repeat the above steps until the optimal solution is found or the maximum number of iterations is reached.

[0100] 5. Result output: The genetic algorithm finally outputs the optimal parking space combination P all and the corresponding maximum comprehensive commuting charging convenience ρ all (P all ). For example, the optimal result of the optimal parking space combination P all is: P all =[P 29 , P7, P 41 , ρ all (P all ) = 1.677.

[0101] 6. Intelligent guidance: According to the optimal result of the genetic algorithm, the system intelligently guides users to the designated parking spaces: User 1: Guide to parking space P 29 , User 2: Guide to parking space P7, User 3: Guide to parking space P 41 .

[0102] To present the beneficial effects of the present invention, this embodiment also provides an average new energy vehicle user satisfaction method. The average new energy vehicle user satisfaction is set according to the following formula: S(P all ) = ∑ i=1 I S Ui (P) / I; S Ui (P) = 1 - (S0(IN j , P) × L(INj , P) + S0(P, OUTk) × L(P, OUTk)).

[0103] Among them, S Ui (P) represents the user satisfaction of the new energy vehicle with user number i; S0(IN j , P) represents the dissatisfaction per unit distance of the new energy vehicle from the entrance of parking lot numbered j to parking space P; S0(P, OUTk) represents the dissatisfaction per unit distance of the new energy vehicle from parking space P to the exit of parking lot numbered k; L(IN j , P) represents the distance of the new energy vehicle from the entrance of parking lot numbered j to parking space P; L(P, OUTk) represents the distance of the new energy vehicle from parking space P to the exit of parking lot numbered k; P all represents the combination of the parking space positions of the new energy vehicles of all users; I represents the total number of users of new energy vehicles that need intelligent guidance for commuting charging and discharging; i represents the user number of the new energy vehicle; S(P all ) represents the average new energy vehicle user satisfaction under the P all combination.

[0104] To evaluate the effect of the intelligent guidance of the three new energy vehicle users in this embodiment, this embodiment uses the average new energy vehicle user satisfaction method to process the three new energy vehicle users under the combination of P all = [P 29 , P7, P 41 . During the processing: S0(IN j , P) is set to 0.001 / m, S0(P, OUTk) is set to 0.0015 / m, L(IN1, P 29 ) is 100m, L(P 29 , OUT2) = 120m, so as to obtain S U1 (P 29 ) = 0.72. Similarly, it can be obtained that S U2 (P7) = 0.68, S U1 (P 41 ) = 0.67, so as to obtain S(P all ) = 0.69.

[0105] Comparative example: This comparative example uses the method of randomly selecting the nearest parking space. After the new energy vehicle with user number i enters the parking lot through the entrance of parking lot numbered j, search for the nearest available parking space P nearest , define P nearest as the center, and the set of available parking spaces within the radius R near is Pnear , randomly select a parking space P from the set P of available parking spaces near as the target parking space for the new energy vehicle of user number i.

[0106] In Embodiment 2, after the new energy vehicle of user number 1 enters the parking lot through the parking lot entrance numbered 1, search for the available parking space P closest to the parking lot entrance numbered 1 nearest which is P 10 , define the set of available parking spaces within a radius R nearest = P 10 = 10m centered on P as the set P near = [P8, P9, P near , P 10 , P 15 , randomly select a parking space P from the parking lot set P near = [P8, P9, P 10 , P 15 as the target parking space for the new energy vehicle of user number 1. Similarly, select P 15 as the target parking space for the new energy vehicle of user number 2, and select P9 as the target parking space for the new energy vehicle of user number 3. 40

[0107] Use the average new energy vehicle user satisfaction method in Embodiment 2 to process the three new energy vehicle users in the combination of P all = [P 15 , P 40 , P9], and the average new energy vehicle user satisfaction S(P all ) = 0.53 of this comparative example is obtained. This result is less than 0.69 obtained in Embodiment 2, indicating that the average new energy vehicle user satisfaction obtained after the intelligent guidance of the new energy vehicle users by the method of the present invention is very good.

[0108] The innovation of the present invention lies in adopting the new energy vehicle user time preference portrait model and the intelligent guidance parking space model, realizing the advantage of comprehensively scheduling the commuting charging and parking of all new energy vehicles, achieving the effect of considering the comprehensive commuting charging convenience and timeliness of all users, and reaching the goal of the optimal comprehensive arrangement of commuting charging and parking.

[0109] The above-described embodiments are only preferred embodiments given to fully illustrate the present invention, and the protection scope of the present invention is not limited thereto. Equivalent substitutions or transformations made by those skilled in the art on the basis of the present invention are all within the protection scope of the present invention. The protection scope of the present invention is subject to the claims.​

Claims

1. A commuting charging and parking intelligent guidance method driven by user time-limit preference, characterized in that The steps include: S1. Collect the behavior data of each new energy vehicle entering and leaving the parking lot in the past few days, and build a parking and charging data set based on the behavior data of all new energy vehicles; S2. Construct a new energy vehicle user time preference portrait model, input the charging and stopping data set into the new energy vehicle user time preference portrait model for training, and obtain a trained new energy vehicle user time preference portrait model; S3. Input the collected behavior data before the current moment into the new energy vehicle user time-efficiency preference portrait model for prediction, and obtain the total commuting parking and charging convenience of each new energy vehicle based on the prediction results and the collected commuting parking area planning data within the preset sliding window time, and obtain the comprehensive commuting parking and charging convenience of all users based on the total commuting parking and charging convenience; S4. Build an intelligent guidance parking space model based on the comprehensive commuting, stopping and charging convenience, use a genetic algorithm to process the intelligent guidance parking space model to obtain the parking space position corresponding to each user's new energy vehicle, and guide each new energy vehicle to enter the corresponding parking space according to the obtained parking space position.

2. The intelligent guiding method for commuting charging and parking driven by user time-limit preference according to claim 1, wherein The S1 is specifically: S11. Collect the behavior data of each new energy vehicle entering and leaving the parking lot every time within a certain number of days in history; S12, counting the probability values ​​of each new energy vehicle leaving the parking lot from each parking lot exit on all days according to the behavior data, and using the probability values ​​of leaving the parking lot from all parking lot exits as probability data; S14, using the behavior data as input data and the corresponding probability data as labels to construct the charging and stopping data of each new energy vehicle; S15. The charging stop data of all new energy vehicles are aggregated to obtain a charging stop data set.

3. According to claim 2, a user time preference driven commuting stop and charge intelligent guidance method is characterized by: The behavior data of each new energy vehicle entering and exiting the parking lot each time in S11 includes the user number of the new energy vehicle, the parking lot entrance number when entering the parking lot, the time period when entering the parking lot entrance, and the parking lot exit number when leaving the parking lot.

4. According to claim 1, a user time preference driven commuting stop and charge intelligent guidance method is characterized by: The new energy vehicle user time preference portrait model in S2 adopts a long short-term memory neural network.

5. The intelligent guiding method for commuting charging and parking driven by user time-limit preference according to claim 1, wherein The S3 is specifically: S31, collecting the behavior data of all new energy vehicles in a preset time period before the current moment, and inputting the behavior data into the trained new energy vehicle user time preference portrait model to predict the probability value of each new energy vehicle leaving the parking lot from each parking lot exit within the preset sliding window time after the current moment; S32, collecting commuting parking area planning data for each new energy vehicle within a preset sliding window time, and obtaining the total commuting parking convenience for each new energy vehicle based on the commuting parking area planning data and the probability value of each new energy vehicle leaving the parking lot from each parking lot exit; S33. Obtain the comprehensive commuting stop-and-charge convenience of all users based on the total commuting stop-and-charge convenience of each new energy vehicle user.

6. A commuting charging and parking intelligent guidance method driven by user time preference as claimed in claim 5, characterized in that: The commuting parking area planning data of each new energy vehicle in S32 includes the user number of the new energy vehicle, the time period when entering the parking lot entrance, the parking lot entrance number when entering the parking lot, the number of the initially estimated parking space, the difficulty value of driving from the parking lot entrance to the parking space, the difficulty value of driving from the parking space to the parking lot exit, the total number of parking lot exits, the power of the new energy vehicle when entering the parking lot entrance, and the density of new energy vehicles in the parking space.

7. A commuting charging and parking intelligent guidance method driven by user time preference as claimed in claim 5, characterized in that: The total commuting charging and parking convenience of each new energy vehicle in S32 is obtained by the following formula: ρ i = 0.5α i + 0.3β i + 0.2γ i α i =1-(D i，INj (P)+∑ k=1 n f i，INj，OUTk ×D i，OUTk (P)) / 2 β i =(1 - D i，INj (P))×(1 - SOC / 100) γ i = 1 - INT(P) Among them, i represents the user number of new energy vehicles; ρ i represents the overall commuting charging convenience of the new energy vehicle with user number i; α i represents the commuting charging convenience based on parking and pick-up of new energy vehicle users; β i represents the commuting charging convenience based on charging of new energy vehicles; γ i represents the commuting charging convenience based on the density of new energy vehicles; both k and j represent indices; D i，INj (P) represents the difficulty value for a new energy vehicle to travel from the entrance of parking lot numbered j to the parking space numbered P; D i，OUTk (P) represents the difficulty value for a new energy vehicle to travel from the parking space numbered P to the exit of parking lot numbered k; n represents the total number of parking lot exits; f i，INj，OUTk represents the probability value for a new energy vehicle to leave the parking lot from the exit of parking lot numbered k on the premise that the new energy vehicle enters the entrance of parking lot numbered j; SOC represents the power of the new energy vehicle when it enters the parking lot entrance; INT(P) represents the density of new energy vehicles in the parking space numbered P.

8. The intelligent guiding method for commuting charging and parking driven by user time-limit preference according to claim 5, characterized in that, S4 specifically is: S41. Binary code the numbers of all parking spaces to obtain the numbers of the parking spaces after binary coding; S42. Determine the constraint conditions according to the numbers of the parking spaces after binary coding, and construct an intelligent guidance parking space model with the comprehensive commuting charging and parking convenience as the fitness function; S43. Use the genetic algorithm to process the intelligent guidance parking space model to obtain the combination of the parking space positions of the new energy vehicles of all users under the maximum comprehensive commuting charging and parking convenience, and obtain the corresponding parking space positions of the new energy vehicles of each user according to the combination of the parking space positions; S44. Guide each new energy vehicle to the corresponding parking space position according to the parking space positions obtained in S43.

9. A commuting charging and parking intelligent guidance method driven by user time preference as claimed in claim 8, characterized in that: The intelligent guidance parking space model in S42 is set according to the following formula: Max(ρ all (P all )) ρ all (P all )=∑ i=1 m ρ i 0<P ibin <P max P ibin ≠P jbin Among them, i represents the user number of new energy vehicles; ρ i represents the overall commuting charging convenience of the new energy vehicle with user number i; m represents the total number of users of new energy vehicles that require intelligent guidance for commuting charging; P all represents the combination of the parking space positions of the new energy vehicles of all users; ρ all (P all ) represents the comprehensive commuting charging convenience of the new energy vehicles of all users under the combination of P all ; Max(ρ all (P all )) represents the maximized fitness function; P ibin represents the binary encoding of the parking space corresponding to the new energy vehicle with user number i; P jbin represents the numbered binary encoding of the parking space corresponding to the new energy vehicle with user number j; P max represents the maximum value among the numbered binary encodings of all parking spaces.

10. An apparatus for implementing the commuting parking and charging intelligent guiding method according to any one of claims 1-9, characterized in that, Including: An information collection module, used to collect the behavior data and commuting parking area planning data of new energy vehicles; An information storage module, used to store the behavior data and commuting parking area planning data of new energy vehicles; A model library module, used to construct and store the new energy vehicle user time preference portrait model and the intelligent guidance parking space model; A calculation module, used to process and calculate the total commuting charging and parking convenience of each new energy vehicle, the comprehensive commuting charging and parking convenience of all users, and the combination of the parking space positions of the new energy vehicles of all users under the maximum comprehensive commuting charging and parking convenience; A visual guidance module, used to guide each new energy vehicle to drive to the corresponding parking space position.

Images