Vehicle intelligent battery swapping navigation method and system based on multi-objective genetic algorithm

Abstract

The present invention discloses a vehicle intelligent battery swap navigation method and system based on a multi-objective genetic algorithm. The method includes: obtaining a power battery loss value and driver behavior data, and performing fuzzy cluster analysis on the driver behavior data; sending a battery swap warning message when the power battery loss value is greater than a preset value or the power battery SOC value is less than a first preset threshold; selecting and entering a battery swap navigation calculation mode, detecting the vehicle's instantaneous vehicle information, and calculating the remaining cruising range based on the SOC value; searching for multiple battery swap stations based on the instantaneous vehicle information and the remaining cruising range, and obtaining battery swap waiting data for each battery swap station and basic road data for the vehicle to each battery swap station, as well as driver behavior data after fuzzy cluster analysis; establishing a multi-objective optimization model to solve for the optimal battery swap station and the optimal path. By analyzing the driver behavior data, the present invention can formulate personalized and efficient battery swap paths and battery swap stations based on the driver's behavior habits.

Description

Technical Field

[0001] The present invention relates to the field of automobile navigation technology, and in particular to a vehicle intelligent battery swapping navigation method and system based on a multi-objective genetic algorithm. Background Art

[0002] To promote green and low-carbon global development and address global climate change, countries around the world are vigorously promoting new energy vehicles. As the "heart" of new energy vehicles, the charging and replacement of power batteries is crucial.

[0003] When users encounter low battery levels in pure electric vehicles, in addition to traditional charging methods, the government encourages battery swapping. Currently, drivers often search for battery swap stations by entering key destinations. There is a lack of an in-vehicle navigation method and system that can monitor battery status in real time, collect and analyze driver behavior data in real time, and develop personalized, efficient battery swap routes and adapt battery swap stations based on driver behavior to meet multiple optimization goals and improve the convenience of battery swapping technology for new energy vehicles. Summary of the Invention

[0004] The present invention aims to address, at least to some extent, one of the technical problems in the related art. To this end, the first object of the present invention is to provide a vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm. By analyzing driver behavior data, a personalized, efficient battery swapping route and adaptive battery swapping station can be developed based on the driver's behavioral habits.

[0005] The second object of the present invention is to provide a vehicle intelligent battery replacement navigation system based on a multi-objective genetic algorithm.

[0006] To achieve the above object, the present invention is implemented through the following technical solutions:

[0007] A vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm, comprising:

[0008] Step S1: acquiring a power battery loss value in real time, collecting driver behavior data, and performing fuzzy cluster analysis on the driver behavior data;

[0009] Step S2: Determine whether the power battery loss value is greater than a preset power battery loss value. If so, execute step S5; if not, execute step S3;

[0010] Step S3: using the open circuit voltage method to detect the initial SOC value of the power battery when the vehicle is started or paused, and using the ampere-hour integration method to detect the SOC value of the power battery in the working state;

[0011] Step S4: Determine whether the SOC value is less than a first preset threshold value. If so, proceed to step S5; otherwise, return to step S1.

[0012] Step S5: Sending battery swap warning information, determining whether the driver has input destination information and whether the driver has accepted the warning information, and selecting a corresponding battery swap navigation calculation mode based on the determination result;

[0013] Step S6: Turn on the selected battery swap navigation calculation mode, detect the vehicle's instantaneous vehicle information in real time, and calculate the remaining cruising range based on the SOC initial value and the SOC value;

[0014] Step S7: Based on the instantaneous vehicle information and the remaining cruising range, a number of battery swap station distribution points are searched and obtained, and the number of battery swap points and the battery swap vehicle capacity of each battery swap station distribution point are obtained. A multi-service station hybrid queuing theory model is established based on the number of battery swap points and the battery swap vehicle capacity of the battery swap station. The average waiting time of the battery swap queue is calculated using the multi-service station hybrid queuing theory model; and basic road data, road congestion, path turning direction data, and driver behavior data after fuzzy cluster analysis are obtained for vehicles traveling to each battery swap station distribution point.

[0015] Based on the basic road data of the vehicle to each battery swap station distribution point, an in-transit travel time prediction model is established, and the vehicle's in-transit travel time is predicted by the in-transit travel time prediction model; the driver's driving behavior habit category is determined based on the path turning direction data and the driver's behavior data after fuzzy cluster analysis, and the total in-transit travel time of the vehicle is determined based on the driving behavior habit category and the vehicle's in-transit travel time, so as to establish a multi-objective optimization model through the total in-transit travel time and the average waiting time in the battery swap queue.

[0016] The multi-objective optimization model is a target optimization model including an objective function and constraints; wherein the objective function includes minimizing the user's total battery swap time and maximizing driving comfort, wherein the user's total battery swap time includes the total on-the-go driving time, the average waiting time in the battery swap queue, and the battery swap operation time, and the battery swap operation time is a constant; driving comfort includes the number of turns on each path; and the constraints include the vehicle's remaining battery capacity constraint, the road traffic congestion constraint, the battery swap station queue length constraint, and the multi-step prediction constraint.

[0017] Step S8: Perform multi-objective genetic optimization calculations through the multi-objective optimization model to obtain the optimal battery swap station and the optimal path.

[0018] Establish the multi-objective optimization model, preset the number of iterations t and the maximum number of iterations t max , and obtain the t-th generation parent population P t; Get the fitness value, and adjust the parent population P according to the fitness value t The population chromosome is selected; the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Perform a crossover operation; perform a mutation operation on the parent population after the crossover to obtain the t-th generation offspring population Q t ; The t-th generation parent population P t and the t-th generation offspring population Q t Merge and use the non-dominated sorting method to obtain the t+1 generation parent population P t+1 ; Get the results of the battery swap station and route allocation; Determine whether the number of iterations is the maximum number of iterations t max If yes, then output the allocation result and get the optimal swap station and optimal path. If no, then return to the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Steps to perform crossover operation.

[0019] Optionally, in step S1, the step of performing fuzzy cluster analysis on the driver behavior data includes:

[0020] Step S11: Obtain n sets of driver behavior data as the initial sample domain, and record the initial sample domain data as x1, x2…, x i , where each sample domain data has m feature vectors, then the feature attribute data of the i-th classification object in the initial sample domain is x i ={x′ i1 ,x′ i2 ,…,x′ im};

[0021] Step S12: Standardize each feature attribute data according to the following formula to obtain the corresponding standardized data:

[0022]

[0023] in, x ij is the standardized data corresponding to each feature attribute data, i = 1, 2, ..., n, j = 1, 2, ..., m;

[0024] Step S13: Create a fuzzy similarity matrix and record x i with x j The similarity coefficient is r ij =R(x i ,x j ), wherein the similarity coefficient satisfies the following conditions:

[0025]

[0026] Among them, x i with x j are two classified data in the initial sample domain, r ij ∈[-1,1], if r ij <0, then And r′ ij ∈[0,1],r′ ij is the similarity coefficient;

[0027] Step S14: Solve the transfer closure function t(R) of the fuzzy similarity coefficient matrix R and obtain t(R)=R n , and take different confidence levels to obtain different clustering results of driver behavior data;

[0028] Step S15: storing the clustering result as a feature vector sample interval, continuing to collect new driver behavior data, and matching the new driver behavior data with the feature vector sample interval. If the match is successful, obtaining the classification result of the driver behavior data; if not, determining that the driver behavior data is irrelevant data;

[0029] Step S16: Continue to collect the accumulated driver behavior data for the preset period, and merge the matching results of the new driver behavior data with the clustering results of the previously stored accumulated driver behavior data, and return to step S11 to repeat the step of performing fuzzy cluster analysis on the driver behavior data.

[0030] Optionally, the driver behavior data includes time consumption data for different turns at intersections and habitual driving speeds at speed-limited sections between speed-limited nodes. The driver behavior data specifically includes: time consumption data for left turns, right turns, and straight-ahead driving at intersections during congested periods, time consumption data for left turns, right turns, and straight-ahead driving at intersections during non-congested periods, driving speeds between speed-limited nodes during congested and non-congested periods on holidays, and driving speeds between speed-limited nodes during congested and non-congested periods on non-holidays.

[0031] Optionally, step S5 includes: when it is determined that the driver has input the destination information and accepted the warning information, selecting the first battery-swap navigation calculation mode; when it is determined that the driver has input the destination information and has not accepted the warning information, selecting the second battery-swap navigation calculation mode; when it is determined that the driver has not input the destination information and has accepted the warning information, selecting the third battery-swap navigation calculation mode; when it is determined that the driver has not input the destination information and has not accepted the warning information, selecting the fourth battery-swap navigation calculation mode; wherein,

[0032] When the first and third battery swap navigation calculation modes are selected, step S6 is directly executed. When the second and fourth battery swap navigation calculation modes are selected, the SOC value of the power battery in the working state is detected, and when the SOC value is less than the second preset threshold, step S6 is forcibly executed.

[0033] Optionally, the step of searching for a plurality of battery swap station distribution points according to the instantaneous vehicle information and the remaining cruising range in step S7 includes:

[0034] When the first and second battery swap navigation calculation modes are selected, the input destination information is obtained, and the instantaneous location of the vehicle is used as the vertex of the sector. The center line of the sector is set along the direction of the destination. Under the condition of meeting the remaining cruising range, the sector coverage area of ​​0° to 180° is searched to obtain several battery swap station distribution points;

[0035] When the third and fourth battery swap navigation calculation modes are selected, the vehicle's instantaneous location is used as the center, and the remaining cruising range condition is met. The circular coverage area of ​​0° to 360° is searched to obtain several battery swap station distribution points.

[0036] Optionally, the average waiting time of the battery swap queue in step S7 is calculated using a multi-service station mixed queuing theory model, and the average waiting time of the battery swap queue is determined using the following formula:

[0037]

[0038] Among them, W q is the average waiting time in the battery replacement queue, L q is the average length, λ e is the effective vehicle arrival rate.

[0039] To achieve the above objectives, the second aspect of the present invention provides a vehicle intelligent battery swapping navigation system based on a multi-objective genetic algorithm, comprising:

[0040] Battery health detection module 10, used to obtain power battery loss value in real time;

[0041] Driver behavior detection module 20, used to collect and send driver behavior data to the database for fuzzy cluster analysis;

[0042] The current detection module 30 is used to detect the initial SOC value of the power battery when the vehicle is started or paused using the open circuit voltage method, and to detect the SOC value of the power battery in the working state using the ampere-hour integration method;

[0043] Warning module 40, used to send battery replacement warning information;

[0044] Satellite positioning module 50, used for detecting the instantaneous vehicle information of the vehicle in real time;

[0045] The battery swap system data platform 60 is used to send data on the number of batteries waiting for swapping, the average waiting time in the battery swap queue, and the number of fully charged batteries at each battery swap station distribution point;

[0046] Road supervision data platform 70, used to send basic road data, road congestion and path turning direction data for vehicles to each battery swap station distribution point;

[0047] Wireless communication module 80, used to transmit data on the number of vehicles waiting for battery swapping, average waiting time in the battery swapping queue, number of fully charged batteries, basic road data for vehicles traveling to each battery swapping station distribution point, road congestion and path turning direction data to microcontroller 90;

[0048] The microcontroller 90 is configured to determine whether the power battery loss value is greater than a preset power battery loss value. If so, the microcontroller controls the early warning module 40 to send a battery replacement early warning message. If not, the microcontroller controls the current detection module 30 to detect the SOC value of the power battery and determine whether the SOC value is less than a first preset threshold value. If so, the microcontroller controls the early warning module 40 to send a battery replacement early warning message. If not, the microcontroller controls the battery health detection module 10 to obtain the power battery loss value in real time. After the early warning module 40 sends the battery replacement early warning message, the microcontroller controls whether the driver has input destination information and whether the driver has accepted the early warning message, and selects a corresponding battery replacement navigation calculation mode based on the determination result.

[0049] After the selected battery swap navigation calculation mode is turned on, the microcontroller 90 is further configured to calculate the remaining cruising range based on the initial SOC value and the SOC value, and search for a number of battery swap station distribution points based on the instantaneous vehicle information and the remaining cruising range, obtain the number of battery swap points within each battery swap station distribution point and the battery swap vehicle capacity of the battery swap station, establish a multi-service station hybrid queuing theory model based on the number of battery swap points within the station and the battery swap vehicle capacity of the battery swap station, and calculate the average waiting time for battery swapping queues using the multi-service station hybrid queuing theory model; and obtain basic road data, road congestion, path turning direction data, and driver behavior data after fuzzy cluster analysis for vehicles traveling to each battery swap station distribution point;

[0050] Based on the basic road data of the vehicle to each battery swap station distribution point, an in-transit travel time prediction model is established, and the vehicle's in-transit travel time is predicted by the in-transit travel time prediction model; the driver's driving behavior habit category is determined based on the path turning direction data and the driver's behavior data after fuzzy cluster analysis, and the total in-transit travel time of the vehicle is determined based on the driving behavior habit category and the vehicle's in-transit travel time, so as to establish a multi-objective optimization model through the total in-transit travel time and the average waiting time in the battery swap queue.

[0051] The multi-objective optimization model is a target optimization model including an objective function and constraints; wherein the objective function includes minimizing the total battery swap time of the user and maximizing driving comfort, wherein the total battery swap time of the user includes the total on-the-go driving time, the average waiting time in the battery swap queue and the battery swap operation time, and the battery swap operation time is a constant; driving comfort includes the number of turns on each path; the constraints include the vehicle's remaining power constraint, the road traffic congestion constraint, the length of the battery swap station queue constraint and the multi-step prediction constraint.

[0052] The multi-objective optimization model is used to perform a multi-objective genetic optimization calculation to obtain the optimal battery swap station and the optimal path. The multi-objective optimization model is established, and the number of iterations t and the maximum number of iterations t are preset. max , and obtain the t-th generation parent population P t ; Get the fitness value, and adjust the parent population P according to the fitness value t The population chromosome is selected; the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Perform a crossover operation; perform a mutation operation on the parent population after the crossover to obtain the t-th generation offspring population Q t ; The t-th generation parent population P t and the t-th generation offspring population Q t Merge and use the non-dominated sorting method to obtain the t+1 generation parent population P t+1 ; Get the results of the battery swap station and route allocation; Determine whether the number of iterations is the maximum number of iterations t max If yes, then output the allocation result and get the optimal swap station and optimal path. If no, then return to the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Steps to perform crossover operation.

[0053] The present invention has at least the following technical effects:

[0054] (1) The present invention makes intelligent improvements to the vehicle's battery swap navigation technology based on the driver's behavioral habits. Specifically, the present invention can collect driver behavior data in real time and train the driver's behavior data based on a fuzzy clustering algorithm, so that the system can continuously classify the driver's behavior data through the "experience" of "autonomous learning", making the classification more accurate; further, when calculating the user's total battery swap time, the system takes into account factors such as the different time spent by the driver at different turns at intersections and the different driving speeds in various speed-limited sections, so as to formulate a personalized and efficient battery swap path for the driver.

[0055] (2) The present invention can simultaneously meet two optimization objectives. Specifically, when establishing a multi-objective model, the present invention takes the minimum total user battery replacement time and the maximum driving comfort as the objective function. Compared with the existing technology that considers a single objective, the present invention not only minimizes the battery replacement time but also improves the driver's driving comfort.

[0056] (3) In the process of solving multi-objective genetic optimization, the present invention designs a unique encoding and crossover method that meets the practical problem of battery replacement path selection for pure electric vehicles. Specifically, the present invention divides the driving path into several sections based on intersections. When performing the encoding and crossover operations of the genetic algorithm, the length of the chromosome is designed to be non-fixed, and during the crossover operation, the chromosomes are paired according to the common gene positions, and corresponding elimination and retention methods are designed. Compared with the traditional genetic algorithm, the present invention is more suitable for solving the multi-objective model and increases the diversity of the model solution.

[0057] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 A flowchart of a vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm provided by one embodiment of the present invention;

[0059] Figure 2 A flowchart of driver behavior data analysis provided by one embodiment of the present invention;

[0060] Figure 3 A flowchart of a multi-objective genetic optimization algorithm provided by one embodiment of the present invention;

[0061] Figure 4 A schematic diagram of a road network provided in one embodiment of the present invention;

[0062] Figure 5 A schematic diagram of chromosome coding provided by one embodiment of the present invention;

[0063] Figure 6 This is a structural block diagram of a vehicle intelligent battery swapping navigation system based on a multi-objective genetic algorithm provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0064] The present embodiment is described in detail below. Examples of the embodiment are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, but are not to be construed as limiting the present invention.

[0065] The following describes the vehicle intelligent battery swapping navigation method and system based on a multi-objective genetic algorithm of this embodiment with reference to the accompanying drawings. It should be noted that the vehicle intelligent battery swapping navigation method and system based on a multi-objective genetic algorithm of this embodiment is applied to new energy pure electric vehicles. It can monitor the power state of the power battery in real time, accurately classify the driver's behavioral habit data, and formulate personalized and efficient battery swapping paths, which can effectively improve the convenience and universality of battery swapping technology for new energy electric vehicles.

[0066] Figure 1 This is a flow chart of a vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm provided by an embodiment of the present invention. Figure 1 As shown, the method includes:

[0067] Step S1: Acquire the power battery loss value in real time, collect driver behavior data, and perform fuzzy cluster analysis on the driver behavior data.

[0068] Specifically, the driver behavior data may include time consumption data for different turns at intersections and customary driving speeds for each speed-limited road section between each speed-limited node. The driver behavior data specifically includes: time consumption data for left turns, right turns, and straight-ahead driving at intersections during congested periods; time consumption data for left turns, right turns, and straight-ahead driving at intersections during non-congested periods; driving speeds between speed-limited nodes during congested and non-congested periods on holidays; and driving speeds between speed-limited nodes during congested and non-congested periods on non-holidays. The details are shown in Table 1 below:

[0069] Table 1 Driver behavior data

[0070]

[0071]

[0072] In step S1, the step of performing fuzzy cluster analysis on the driver behavior data includes:

[0073] Step S11: Obtain n sets of driver behavior data as the initial sample domain, and record the initial sample domain data as x1, x2…, x i , where each sample domain data has m feature vectors, then the feature attribute data of the i-th classification object in the initial sample domain is x i ={x′ i1 ,x′ i2 ,…,x′ im}.

[0074] The vehicle intelligent battery swap navigation method in this embodiment is applied to the vehicle intelligent battery swap navigation system. Specifically, the background server of the battery swap navigation system retrieves the latest n sets of driver behavior data from the database as a set of initial sample domains, which include the time consumed by the driver when turning left, turning right, and going straight during congested periods, non-congested periods, and holidays, and the habitual driving speed on each speed-limited road section, and then extracts the same features of the samples to form a feature vector. The initial sample domain data is denoted as x1, x2…, x i , where each sample domain data has m feature vectors, then the feature attribute data of the i-th classification object in the initial sample domain is x i ={x′ i1 ,x′ i2 ,…,x′ im}.

[0075] Step S12: Standardize each feature attribute data according to the following formula to obtain corresponding standardized data:

[0076]

[0077] in, x ij is the standardized data corresponding to each feature attribute data, i = 1, 2, ..., n, j = 1, 2, ..., m.

[0078] Step S13: Create a fuzzy similarity matrix and record x i with x j The similarity coefficient is r ij =R(x i ,x j ), where the similarity coefficient satisfies the following conditions:

[0079]

[0080] Among them, x i with x j are two classified data in the initial sample domain, r ij ∈[-1,1], if r ij <0, then And r′ ij ∈[0,1],r′ ij is the similarity coefficient.

[0081] Step S14: Solve the transfer closure function t(R) of the fuzzy similarity coefficient matrix R and obtain t(R)=R n , and take different confidence levels to obtain different clustering results of driver behavior data. Among them, the confidence level ranges from 0 to 1.

[0082] Step S15: Store the clustering results and use them as feature vector sample intervals, continue to collect new driver behavior data, and match the new driver behavior data with the feature vector sample intervals. If the match is successful, obtain the classification result of the driver behavior data. If not, determine that the driver behavior data is irrelevant data.

[0083] Specifically, such as Figure 2 As shown, the battery swap navigation system server first stores the initial clustering results and returns the results to the battery swap navigation system as an updated feature vector sample interval. The battery swap navigation system continues to collect data and match it with the feature vector sample interval. If the match is successful, the classification result corresponding to the data is returned. If there is no match, the data is judged to be irrelevant data and is eliminated.

[0084] Step S16: Continue to collect the accumulated driver behavior data for the preset period, and merge the matching results of the new driver behavior data with the clustering results of the previously stored accumulated driver behavior data, and return to step S11 to repeat the step of performing fuzzy cluster analysis on the driver behavior data.

[0085] Specifically, the battery swap navigation system continues to collect cumulative driver behavior data for a preset period of time, and the battery swap navigation system server stores multiple groups of clustering results. After a period of accumulation, the new matching results can be merged with the stored multiple groups of clustering results, and then return to step S11 to repeat the fuzzy clustering process.

[0086] Step S2: Determine whether the power battery loss value is greater than a preset power battery loss value. If so, execute step S5; if not, execute step S3.

[0087] In this embodiment, the specific value of the preset power battery loss value can be determined according to the model of the power battery used and the model of the vehicle.

[0088] Step S3: using the open circuit voltage method to detect the initial SOC value of the power battery when the vehicle is started or paused, and using the ampere-hour integration method to detect the SOC value of the power battery in the working state.

[0089] In this embodiment, the SOC (State Of Charge) value of the power battery is detected by combining the open circuit voltage method and the ampere-hour integration method. The initial value of the power battery SOC refers to the power level of the power battery when the battery is just started or stopped.

[0090] Step S4: Determine whether the SOC value is less than a first preset threshold value. If so, execute step S5; otherwise, return to step S1.

[0091] In this embodiment, the value of the first preset threshold may be within the range of 30%-50% of the full charge of the power battery, but its specific value is not specifically limited here and can be set according to actual conditions.

[0092] Step S5: Send battery swap warning information, determine whether the driver has entered the destination information and whether the driver has accepted the warning information, and select the corresponding battery swap navigation calculation mode based on the judgment result.

[0093] The step S5 includes: when it is determined that the driver has input the destination information and accepted the warning information, selecting the first battery-swap navigation calculation mode; when it is determined that the driver has input the destination information and has not accepted the warning information, selecting the second battery-swap navigation calculation mode; when it is determined that the driver has not input the destination information and accepted the warning information, selecting the third battery-swap navigation calculation mode; when it is determined that the driver has not input the destination information and has not accepted the warning information, selecting the fourth battery-swap navigation calculation mode; wherein, when the first and third battery-swap navigation calculation modes are selected, step S6 is directly executed; when the second and fourth battery-swap navigation calculation modes are selected, the SOC value of the power battery in the working state is detected, and when the SOC value is less than the second preset threshold value, step S6 is forcibly executed.

[0094] Specifically, the working principles of the first to fourth battery swap navigation calculation modes in this embodiment are similar. The specific working steps of the four working modes are all to execute steps S6-S8. The difference is that for the first battery swap navigation calculation mode and the third battery swap navigation calculation mode, when the battery swap navigation system issues an early warning, if the driver accepts the early warning, he will directly enter step S6; for the second battery swap navigation calculation mode and the fourth battery swap navigation calculation mode, when the battery swap navigation system issues an early warning, if the driver does not respond for many times, he will continue to drive until the power battery power is less than the second preset threshold value, and the battery swap navigation system will forcibly start the battery swap navigation calculation mode and enter step S6.

[0095] Among them, when entering the first and second battery swap navigation calculation modes, the 0°~180° fan-shaped search method is used to obtain the battery swap station distribution points. When entering the third and fourth battery swap navigation calculation modes, the 360° circular search method is used to obtain the battery swap station distribution points.

[0096] Step S6: Turn on the selected battery swap navigation calculation mode, detect the vehicle's instantaneous vehicle information in real time, and calculate the remaining cruising range based on the SOC initial value and SOC value.

[0097] The instantaneous vehicle information of the vehicle includes the instantaneous speed, instantaneous position, and instantaneous direction of the vehicle. In this embodiment, the instantaneous remaining cruising range can also be calculated based on the SOC initial value and the SOC value.

[0098] Step S7: According to the instantaneous vehicle information and the remaining cruising range, several battery swap station distribution points are searched and obtained, and the number of battery swap points and the battery swap vehicle capacity of each battery swap station distribution point are obtained. According to the number of battery swap points and the battery swap vehicle capacity of the battery swap station, a multi-service station mixed queuing theory model is established, and the average waiting time of the battery swap queue is calculated through the multi-service station mixed queuing theory model; and the basic road data, road congestion, path turning direction data and driver behavior data after fuzzy clustering analysis for vehicles to each battery swap station distribution point are obtained.

[0099] The step of searching for a number of battery swap station distribution points based on the instantaneous vehicle information and the remaining cruising range in step S7 includes: when selecting the first and second battery swap navigation calculation modes, obtaining the input destination information, and using the vehicle's instantaneous location as the sector vertex, setting the sector center line along the destination direction, and searching the sector coverage area of ​​0° to 180° while satisfying the remaining cruising range condition to obtain a number of battery swap station distribution points; when selecting the third and fourth battery swap navigation calculation modes, using the vehicle's instantaneous location as the center, and satisfying the remaining cruising range condition, searching the circular coverage area of ​​0° to 360° to obtain a number of battery swap station distribution points.

[0100] The average waiting time of the battery swap queue in step S7 is calculated using a multi-service station mixed queuing theory model, and the following formula is used to determine the average waiting time of the battery swap queue:

[0101]

[0102] Among them, W q is the average waiting time in the battery replacement queue, L q is the average length, λ e is the effective vehicle arrival rate.

[0103] Specifically, the calculation of the average waiting time for battery swapping is in accordance with the M / M / s / k multi-service station mixed queuing theory model, in which the average time interval of new energy vehicles arriving at the battery swapping station obeys a negative exponential distribution with an average arrival rate of λ, and the average battery swapping time of new energy vehicles obeys a negative exponential distribution with a service rate of μ, where s is the number of battery swapping points in the station, k is the battery swapping capacity of the battery swapping station, ρ represents the busyness of the battery swapping station system, and defines The specific definition process is as follows:

[0104] Set the value conditions of parameters λ and μ:

[0105]

[0106] Assume the probability distribution p of the system equilibrium state n , system idle probability p0 and potential unit service intensity ρS :

[0107]

[0108] By the stationary distribution p n , n=0,1,2……k, set the average queue length L s , Average Captain L q :

[0109]

[0110] Derived L s =L q +ρ(1+p k ),in is the customer loss rate, specifically, the proportion of customers who are unable to enter the battery swap station among all customers;

[0111] Furthermore, define the effective vehicle arrival rate λ e , average length of stay W s and the average waiting time W q :

[0112] λ e =λ(1-p k ) (10)

[0113]

[0114] Based on the basic road data of the vehicle to each battery swap station distribution point, an en route travel time prediction model is established, and the vehicle's en route travel time is predicted by the en route travel time prediction model; the driver's driving behavior habit category is determined based on the path turning direction data and the driver's behavior data after fuzzy clustering analysis, and the total en route travel time of the vehicle is determined based on the driving behavior habit category and the vehicle's en route travel time, so as to establish a multi-objective optimization model through the total en route travel time and the average waiting time in the battery swap queue.

[0115] The in-transit travel time prediction model is established using the principle of a multi-step prediction method. The principle of the multi-step prediction method is to combine the actual traffic flow data of the starting period and the historical traffic flow data of the subsequent periods to predict the travel time consumption. The specific definition process is as follows:

[0116] Processing road network vehicle history time:

[0117]

[0118] in, is the historical average travel time of vehicles passing through the i-th road section in period t, is the real-time travel time of the vehicle passing through the i-th road section in period t, is the historical travel time of the vehicle passing through the i-th road section in the t period of the previous cycle. This value can be retrieved from the historical traffic database. j is the capacity of this type of historical data. β is the weight coefficient, which is 0.4. type is the historical data type value. The value of type is 0, 1, and 2, representing holidays, weekends, and weekdays, respectively.

[0119] Furthermore, a multi-step prediction model for the travel time of segmented paths from starting point A to B in the road network is established:

[0120]

[0121] in, is the multi-step predicted value of the total travel time of the vehicle starting from A and traveling along the AB path during period t, is the predicted travel time of the vehicle starting at time t on the i-th section of the AB path, Δt i-1 It is the time consumed by the vehicle starting from time period t and traveling along the first i-1 sections of the AB path.

[0122] Since there is a certain time interval between data collection and uploading, the time interval can be set as a. Then a multi-step time prediction model based on time interval is established:

[0123]

[0124] in, is the predicted travel time of the vehicle through the i-th road segment of the AB path during the period ((k-1)a,ka], Δ i-1 is the time period that the vehicle takes to pass through the first i-1 road segments of the AB path, and n1 is the total number of segments divided into the AB path.

[0125] After dividing the time periods into time intervals, the historical average travel time of vehicles passing through road section i in each time period is obtained by combining formula (12). Combined with the real-time traffic network data of the current time period, formulas (15) and (16) are further obtained through exponential smoothing:

[0126]

[0127] in, For vehicles at k+α i-1 The historical average travel time of the i-th road segment of the AB path during the period, For vehicles at k+Δ i-1The average travel time prediction value of the i-th road segment through the AB path during the period, α(Δ i-1 ) is the weight coefficient, and H is the number of potential changes.

[0128] Furthermore, by substituting formula (15) into formula (16) and combining it with formula (14), we can obtain the multi-step prediction value of the total travel time of the vehicle starting from A on the AB path during time period t, as shown in the following formula (17):

[0129]

[0130] It should be noted that the time drivers spend at each intersection varies depending on factors such as the turning direction. Generally, the time spent on turning left and making a U-turn is twice that of going straight, and the time spent on turning right is 1.5 times that of going straight.

[0131] In addition, the turning time consumption is also affected by the driving behavior and habits of different drivers, the type of intersection, and holidays. In this embodiment, the driver's driving behavior habit category is also considered to determine the turning time consumption according to the driver's driving behavior habit category. The intersection types include congested intersections and non-congested intersections. The driver's behavior and habits refer to the different time consumption of different drivers for different turns, and congested intersections and holidays will increase the turning time. Therefore, the above-mentioned fuzzy clustering method is combined to classify the driver's behavior habits at the intersection, and the classification results are compared with the Addition:

[0132]

[0133] Among them, T X is the total travel time, D is the driver's straight travel time, G1 is the number of straight travel nodes, G2 is the right turn time, G3 is the left turn time, γ is the ratio of the right turn time to the straight travel time, and η is the ratio of the left turn time to the straight travel time. It should be noted that γ and η are affected by the classification results of the fuzzy clustering algorithm and vary.

[0134] Furthermore, since drivers have different habitual driving speeds on different speed-limited sections, the drivers' habitual driving speeds on different types of sections can be classified according to the above fuzzy clustering method, and the classified habitual speeds can be used in the above travel time prediction process.

[0135] It should be noted that when retrieving road network traffic flow, if the predicted section is during holiday peak hours or weekday peak hours, the driving speed will use the traffic flow data in its historical database instead of the driver's habitual speed. The following formula is generally used to calculate the time variable:

[0136]

[0137] Among them, T represents the time variable between two nodes, R ij is the length of the path between two nodes ij, v ij is the vehicle speed between two nodes ij.

[0138] Step S8: Perform a multi-objective genetic optimization calculation through a multi-objective optimization model to obtain the optimal battery swap station and the optimal route.

[0139] Among them, the multi-objective optimization model is a target optimization model that includes objective functions and constraints; the objective functions include minimizing the user's total battery replacement time and maximizing driving comfort, among which the user's total battery replacement time includes the total on-the-go driving time, the average waiting time in the battery replacement queue and the battery replacement operation time, and the battery replacement operation time is a constant; driving comfort includes the number of turns on each path; the constraints include the vehicle's remaining power constraint, the road traffic congestion constraint, the battery replacement station queue length constraint and the multi-step prediction constraint.

[0140] like Figure 3 As shown, the step S8 includes: establishing a multi-objective optimization model, presetting the number of iterations t and the maximum number of iterations t max , and obtain the t-th generation parent population P t ; Get the fitness value, and adjust the parent population P according to the fitness value t The population chromosome is selected; the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Perform a crossover operation; perform a mutation operation on the parent population after the crossover to obtain the t-th generation offspring population Q t ; The t-th generation parent population P t and the t-th generation offspring population Q t Merge and use the non-dominated sorting method to obtain the t+1 generation parent population P t+1 ; Get the results of the battery swap station and route allocation; Determine whether the number of iterations is the maximum number of iterations t max If yes, then output the allocation result and get the optimal swap station and optimal path. If no, then return the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Steps to perform crossover operation.

[0141] Specifically, step S8 may include:

[0142] Step S81: Establish a multi-objective optimization model:

[0143] A multi-objective optimization model was established, consisting of an objective function and constraints. The objective function consisted of minimizing the total battery swapping time f1 and maximizing driving comfort f2. The total battery swapping time consists of three components: total travel time, average waiting time in the swapping queue, and swapping operation time. Comfort refers to the number of turns per route; fewer turns result in higher driving comfort. Constraints included the vehicle's remaining battery charge constraint g1, the road traffic congestion constraint g2, the length of the swap station queue g3, and the multi-step prediction constraint g4.

[0144] Among them, the objective function is:

[0145]

[0146] The constraints are:

[0147] The vehicle remaining power constraint (i.e., search range constraint) g1 is defined as:

[0148] E s >∑d ij ×e (23)

[0149] The road traffic congestion constraint g2 is defined as:

[0150] 0 <S<0.75 (24)

[0151] The queue length constraint g3 at the battery swap station is defined as:

[0152] L q +s≤k (25)

[0153] The prediction constraint g4 is defined as:

[0154] Δ i-1 >1 (26)

[0155] Among them, E s The remaining battery power of the vehicle during navigation search, d ij is the distance between nodes on the path, e is the vehicle energy consumption coefficient; S is the saturation, and S = V / C, C = L × c × N car , V is the actual traffic volume of the intersection entrance lane, C is the traffic capacity of the intersection entrance lane, L is the lane mileage, c is the lane capacity, N car is the number of lanes; L q is the average length of the team, s is the number of electric potential swaps within the station, and k is the capacity of electric vehicles swapped at the station; Δ i-1 is the time period that the vehicle spends passing through the first i-1 sections of the AB path, N is the number of intersections on each path leading to the battery swap station within the search range, n2 is the number of turns at the intersection, and T change The unit battery replacement operation time is generally a fixed constant by default.

[0156] Step S82: Set the number of iterations and the maximum number of iterations:

[0157] In this embodiment, the number of iterations can be set to t, and t=0 can be initialized, and the maximum number of iterations can be set to t max .

[0158] Step S83: Get the t-th generation parent population (primary generation population) P t :

[0159] Wherein, step S83 includes:

[0160] Step S831: Assume that the number of available combinations of battery swap stations within the range of the remaining driving range and all feasible roads leading to the battery swap stations is M. A matching result, i.e., the matched battery swap station and driving path, is represented by a chromosome, and the number of chromosomes is M. M chromosomes constitute the initial parent population.

[0161] Step S832: Figure 4 As shown, the selected path consists of nodes and arcs. Let G = {g i |i=1,2,3…,n} is a node set, representing the endpoints of the road segment, and the length of the chromosome gene bit is N1; the chromosome is encoded, such as Figure 5 As shown, the first gene bit is the starting point, the second gene bit is randomly selected from other nodes connected to the starting point, and the selected node is deleted from the node set to prevent duplication. This is repeated when the encoding reaches the destination battery swap station node. The last gene bit is the battery swap station code.

[0162] It should be noted that, in this embodiment, the chromosome coding complies with the chromosome coding regulations, that is, the chromosome genes are not allowed to have repeated gene sites, and the chromosome lengths are not completely the same.

[0163] Step S84: performing a selection operation on the chromosome, specifically including:

[0164] Step S841: According to the parent population P t The objective function value of each chromosome in the quantifier evaluates its individual fitness, and the elite selection strategy is adopted to select P t The individual with the highest fitness in the population is directly retained to the next generation. The fitness of chromosome i, fitness(i), is calculated according to the following formula:

[0165]

[0166] Among them, fitness k (i) is the k-th fitness of chromosome i, f k (i) is the k-th objective function value of chromosome i.

[0167] Step S842: Use the roulette strategy to select the remaining individuals. The probability of individual i being selected is calculated according to the following formula:

[0168]

[0169]

[0170] where p i is the probability that chromosome i is selected, q i is the cumulative fitness value, q N is the cumulative probability.

[0171] Furthermore, a random number between 0 and 1 is randomly generated, and the individual represented by the right endpoint of the cumulative probability interval where the random number is located is extracted, and a sampling method with replacement is adopted, and this step is repeated until the number of chromosomes extracted is M.

[0172] Step S85: performing a crossover operation on the chromosome population, specifically including:

[0173] Step S851: Figure 3 As shown, the parent population P to be extracted t The chromosomes with common genes form a pair. If the number of chromosomes with common genes is odd, the chromosome with the smallest fitness will be eliminated. If there is no common point and the chromosome with a fitness value less than the fitness mean will be directly eliminated, and the chromosome with a fitness value greater than the mean will be retained without crossover operation.

[0174] Specifically, chromosomes are paired according to the common gene pairing method such as Figure 4 As shown, the path A-6-13-20-27-28-29-36-37-C and the path A-1-2-8-15-22-29-30-31-32-B are paired, node A is the instantaneous position when the path search starts, node B and node C are the positions of the battery swap station, and node 29 is the common gene on the chromosome. In the embodiment of the present invention, pairing can only be achieved when two chromosomes have common gene positions, and there is a possibility of crossover.

[0175] Step S852: Different from the traditional genetic algorithm, a random number between 0 and 1 is randomly generated. If the random number is less than the set crossover probability P c , randomly select a common gene as e i , the two chromosomes in each pair of chromosomes are crossed at the gene position to obtain the crossover chromosomes, otherwise the two chromosomes remain unchanged. The crossover chromosomes and the unchanged chromosomes constitute the parent population after the crossover. The crossover probability is calculated according to the following formula:

[0176]

[0177] Among them, a1 is a 0-1 constant, a2 is a 0-1 constant, is the maximum value of fitness, is the average fitness of the group, β k is the weight coefficient, m is the number of objective functions.

[0178] Step S86: Perform mutation operation on the parent population after crossover to obtain the t-th generation offspring population Q t , specifically including:

[0179] Step S861: Calculate the following formula for each individual in the parent population after crossover:

[0180]

[0181] Among them, P m is the mutation probability, f max is the maximum fitness of the group, f avg is the average fitness value of the group, f′ is the fitness value of the mutant individual, k1∈[0.001,0.01], k2∈[0.01,0.1] are constants in the interval.

[0182] Step S862: Generate a random number between 0 and 1 for the individual in the parent population after crossover. If the random number is less than P m , randomly select a gene site as the mutation gene point, then the genes from the first node to the mutation point remain unchanged, and the genes after the mutation point are randomly selected from the adjacent reachable nodes, and so on, until the destination node; if the random number is greater than P m , the chromosome remains unchanged, and the mutated chromosome is obtained, forming the t-th generation offspring population Q t .

[0183] Step S87: Obtain the t+1th generation parent population P t+1 , this step specifically includes:

[0184] Step S871: The t-th generation parent population P t and the t-th generation offspring population Q t Merge to get the t-th generation merged population R t , R t =P t +Q t .

[0185] Step S872: R t The chromosome individuals in R are sorted non-dominatedly, the objective function value of each chromosome is calculated, the dominance and non-domination relationship between individuals is compared, all non-dominated individuals are grouped into the first level non-dominated layer F1, and all chromosomes in F1 are removed from R tRemoved.

[0186] Furthermore, for R t The remaining chromosomes in step S871 are cycled until R t The number of chromosomes in is 0, and q non-dominated layers F1,…,F a ,…,F q .

[0187] Step S873: Set g=1.

[0188] Step S874: Calculate the chromosomes n1′,…,n in the non-dominated layer q ′, judge n1′+…+n g ′=M is established, if so, the chromosomes of the first a layers are taken as the t+1 generation parent population P t+1 The chromosomes of t+1 generation parent population are formed, otherwise step S875 is executed.

[0189] Step S875: Determine n1′+…+n g ' <M,n1'+…+n g '+n g+1 '> Is M satisfied? If so, calculate F a+1 Crowding distance of chromosomes in a layer in is the crowding distance of the i-th chromosome, and sorts them in descending order, selects the first M-(n1'+…+n q ') chromosomes corresponding to the crowding distances, and F1,…F a All chromosomes of the layer are used as the t+1 generation parent population P t+1 The chromosomes of t+1 generation parent population, where the crowding distance The calculation method is as follows:

[0190]

[0191] Where m is the number of objective functions, f k (i) is the kth objective function value of the i-th chromosome, f k (i+1) is the kth objective function value of the i+1th chromosome, f k (i-1) is the k-th objective function value of chromosome i-1, is the maximum value of the kth objective function, is the minimum value of the kth objective function.

[0192] Step S88: Battery swap station and route allocation results.

[0193] Step S89: Determine whether the number of iterations is the maximum number of iterations tmax If yes, then output the allocation result and get the optimal swap station and optimal path, that is, the Pareto solution set. If no, then return the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Steps to perform crossover operation.

[0194] Furthermore, after obtaining the Pareto solution set, if the number of results in the solution set is greater than or equal to 3, the battery swap navigation system can randomly select three matching results to display.

[0195] Figure 6 This is a structural block diagram of a vehicle intelligent battery swapping navigation system based on a multi-objective genetic algorithm provided by the present invention. Figure 6 As shown, the vehicle intelligent battery replacement navigation system 1 based on multi-objective genetic algorithm includes: a battery health detection module 10, a driver behavior detection module 20, a current detection module 30, an early warning module 40, a satellite positioning module 50, a battery replacement system data platform 60, a road supervision data platform 70, a wireless communication module 80 and a microcontroller 90.

[0196] Among them, the battery health detection module 10, the driver behavior detection module 20, the current detection module 30, the early warning module 40, the satellite positioning module 50, the wireless communication module 80 and the microcontroller 90 are all vehicle-mounted devices. The battery swap system data platform 60 is an external data server for storing and processing dynamic data of battery swap stations within the supervision scope, which is matched with the battery swap navigation system. The road supervision data platform 70 is a data platform provided by the traffic management data platform of the Ministry of Public Security and related map software cooperation enterprises.

[0197] Specifically, the battery health detection module 10 is used to obtain the power battery loss value in real time, the driver behavior detection module 20 is used to collect and send driver behavior data to the database for fuzzy clustering analysis, the current detection module 30 is used to use the open circuit voltage method to detect the initial SOC value of the power battery when the vehicle starts or stops working, and use the ampere-hour integration method to detect the SOC value of the power battery in the working state, the early warning module 40 is used to send battery replacement early warning information, the satellite positioning module 50 is used to detect the instantaneous vehicle information of the vehicle in real time, the battery replacement system data platform 60 is used to send the number of battery replacement waiting points, the average waiting time of the battery replacement queue and the number of fully charged batteries at each battery replacement station distribution point, the road supervision data platform 70 is used to send the basic road data, road congestion and path turning direction data of vehicles going to each battery replacement station distribution point, the wireless communication module 80 is used to transmit the number of battery replacement waiting points, the average waiting time of the battery replacement queue, the number of fully charged batteries, the basic road data, road congestion and path turning direction data of vehicles going to each battery replacement station distribution point to the microcontroller 90.

[0198] In this embodiment, the microcontroller 90 is used to determine whether the power battery loss value is greater than a preset power battery loss value. If so, the microcontroller 90 controls the warning module 40 to send a battery swap warning message and displays the battery swap warning message on the screen of the vehicle center console. If not, the microcontroller 90 controls the current detection module 30 to detect the SOC value of the power battery and determines whether the SOC value is less than a first preset threshold value. If so, the microcontroller 90 controls the warning module 40 to send a battery swap warning message. If not, the microcontroller 90 returns to control the battery health detection module 10 to obtain the power battery loss value in real time. Furthermore, after the warning module 40 sends the battery swap warning message, the microcontroller 90 determines whether the driver has entered the destination information and whether the driver has accepted the warning message, and selects the corresponding battery swap navigation calculation mode based on the determination result.

[0199] After the selected battery swap navigation calculation mode is activated, the microcontroller 90 is further configured to calculate the remaining cruising range based on the initial SOC value and the SOC value, and to search for a number of battery swap station distribution points based on the instantaneous vehicle information and the remaining cruising range, obtain the number of battery swap points within each battery swap station distribution point and the battery swap vehicle capacity of the battery swap station, establish a multi-service station hybrid queuing theory model based on the number of battery swap points within the station and the battery swap vehicle capacity of the battery swap station, and calculate the average waiting time for battery swapping queues using the multi-service station hybrid queuing theory model; and obtain basic road data, road congestion, path turning direction data, and driver behavior data after fuzzy cluster analysis for vehicles traveling to each battery swap station distribution point;

[0200] Based on the basic road data of the vehicle to each battery swap station distribution point, an in-transit travel time prediction model is established, and the vehicle's in-transit travel time is predicted by the in-transit travel time prediction model; the driver's driving behavior habit category is determined based on the path turning direction data and the driver's behavior data after fuzzy cluster analysis, and the total in-transit travel time of the vehicle is determined based on the driving behavior habit category and the vehicle's in-transit travel time, so as to establish a multi-objective optimization model through the total in-transit travel time and the average waiting time in the battery swap queue.

[0201] The multi-objective optimization model is a target optimization model including an objective function and constraints; wherein the objective function includes minimizing the total battery swap time of the user and maximizing driving comfort, wherein the total battery swap time of the user includes the total on-the-go driving time, the average waiting time in the battery swap queue and the battery swap operation time, and the battery swap operation time is a constant; driving comfort includes the number of turns on each path; the constraints include the vehicle's remaining power constraint, the road traffic congestion constraint, the length of the battery swap station queue constraint and the multi-step prediction constraint.

[0202] The multi-objective optimization model is used to perform a multi-objective genetic optimization calculation to obtain the optimal battery swap station and the optimal path. The multi-objective optimization model is established, and the number of iterations t and the maximum number of iterations t are preset.max , and obtain the t-th generation parent population P t ; Get the fitness value, and adjust the parent population P according to the fitness value t The population chromosome is selected; the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Perform a crossover operation; perform a mutation operation on the parent population after the crossover to obtain the t-th generation offspring population Q t ; The t-th generation parent population P t and the t-th generation offspring population Q t Merge and use the non-dominated sorting method to obtain the t+1 generation parent population P t+1 ; Get the results of the battery swap station and route allocation; Determine whether the number of iterations is the maximum number of iterations t max If yes, then output the allocation result and get the optimal swap station and optimal path. If no, then return to the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Steps to perform crossover operation.

[0203] It should be noted that the vehicle intelligent battery swap navigation system based on multi-objective genetic algorithm in this embodiment adopts the above-mentioned vehicle intelligent battery swap navigation method based on multi-objective genetic algorithm. Its specific implementation method can refer to the specific implementation method of the above-mentioned vehicle intelligent battery swap navigation method based on multi-objective genetic algorithm. To avoid redundancy, it will not be repeated here.

[0204] In summary, the present invention can collect driver behavior data in real time and train the driver's behavior data based on a fuzzy clustering algorithm, so that the system can continuously classify the driver's behavior data through the "experience" of "autonomous learning", making its classification more accurate. In addition, when calculating the user's total battery replacement time, the present invention takes into account factors such as the different time consumption of the driver in different turns at intersections and the different habitual driving speeds in various speed-limited sections, and formulates a personalized and efficient battery replacement path for the driver; in addition, when establishing a multi-objective model, the present invention takes the minimum user's total battery replacement time and the maximum driving comfort as the objective function. Compared with the existing technology that considers a single objective, the present invention not only shortens the battery replacement time, but also improves the driver's driving comfort; and, the present invention divides the driving path into several sections according to the intersection. When performing the encoding and crossover operations of the genetic algorithm, the length of the designed chromosome is not fixed, and during the crossover operation, the chromosomes are paired according to the common gene positions, and corresponding elimination and retention methods are designed. Compared with the traditional genetic algorithm, it is more suitable for solving the multi-objective model of the present invention and increases the diversity of the model solution.

[0205] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0206] Although the present invention has been described in detail through the above preferred embodiments, it should be understood that the above description is not intended to limit the present invention. After reading the above description, various modifications and substitutions of the present invention will become apparent to those skilled in the art. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm is characterized by: include: Step S1: acquiring a power battery loss value in real time, collecting driver behavior data, and performing fuzzy cluster analysis on the driver behavior data; Step S2: Determine whether the power battery loss value is greater than a preset power battery loss value. If so, execute step S5; if not, execute step S3; Step S3: using the open circuit voltage method to detect the initial SOC value of the power battery when the vehicle is started or paused, and using the ampere-hour integration method to detect the SOC value of the power battery in the working state; Step S4: Determine whether the SOC value is less than a first preset threshold value. If so, proceed to step S5; otherwise, return to step S1. Step S5: Sending battery swap warning information, determining whether the driver has input destination information and whether the driver has accepted the warning information, and selecting a corresponding battery swap navigation calculation mode based on the determination result; Step S6: Turn on the selected battery swap navigation calculation mode, detect the vehicle's instantaneous vehicle information in real time, and calculate the remaining cruising range based on the SOC initial value and the SOC value; Step S7: Searching for a number of battery swap station distribution points based on the instantaneous vehicle information and the remaining cruising range, obtaining the number of battery swap points within each battery swap station distribution point and the battery swap vehicle capacity of the battery swap station, establishing a multi-service station hybrid queuing theory model based on the number of battery swap points within the station and the battery swap vehicle capacity of the battery swap station, and calculating the average waiting time of the battery swap queue using the multi-service station hybrid queuing theory model; It also obtains basic road data, road congestion, path turning direction data, and driver behavior data after fuzzy cluster analysis for vehicles traveling to each battery swap station distribution point; Based on the basic road data of the vehicle to each battery swap station distribution point, a travel time prediction model is established, and the vehicle travel time is predicted by the travel time prediction model; The driver's driving behavior habit category is determined based on the path turning direction data and the driver behavior data after fuzzy cluster analysis, and the total en route driving time of the vehicle is determined based on the driving behavior habit category and the vehicle's en route driving time, so as to establish a multi-objective optimization model based on the total en route driving time and the average waiting time in the battery swap queue. The multi-objective optimization model is a target optimization model including an objective function and constraints; wherein the objective function includes minimizing the user's total battery swap time and maximizing driving comfort, wherein the user's total battery swap time includes the total on-the-go driving time, the average waiting time in the battery swap queue, and the battery swap operation time, and the battery swap operation time is a constant; driving comfort includes the number of turns on each path; and the constraints include the vehicle's remaining battery capacity constraint, the road traffic congestion constraint, the battery swap station queue length constraint, and the multi-step prediction constraint. Step S8: Perform multi-objective genetic optimization calculations through the multi-objective optimization model to obtain the optimal battery swap station and the optimal path. Establish the multi-objective optimization model, preset the number of iterations t and the maximum number of iterations t max , and obtain the t-th generation parent population P t ; Get the fitness value, and adjust the parent population P according to the fitness value t The population chromosome is selected; the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Perform a crossover operation; perform a mutation operation on the parent population after the crossover to obtain the t-th generation offspring population Q t ; The t-th generation parent population P t and the t-th generation offspring population Q t Merge and use the non-dominated sorting method to obtain the t+1 generation parent population P t+1 ; Get the results of the battery swap station and route allocation; Determine whether the number of iterations is the maximum number of iterations t max If yes, then output the allocation result and get the optimal swap station and optimal path. If no, then return to the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Steps to perform crossover operation.

2. The vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm according to claim 1, characterized in that: In step S1, the step of performing fuzzy cluster analysis on the driver behavior data includes: Step S11: Obtain n sets of driver behavior data as the initial sample domain, and record the initial sample domain data as x1, x2…, x i , where each sample domain data has m feature vectors, then the feature attribute data of the i-th classification object in the initial sample domain is x i ={x' i1 ,x' i2 ,…,x' im }; Step S12: Standardize each feature attribute data according to the following formula to obtain corresponding standardized data: in, x ij is the standardized data corresponding to each feature attribute data, i = 1, 2, ..., n, j = 1, 2, ..., m; Step S13: Create a fuzzy similarity matrix and record x i with x j The similarity coefficient is r ij =R(x i ,x j ), wherein the similarity coefficient satisfies the following conditions: Among them, x i with x j are two classified data in the initial sample domain, If r ij <0, then And there is r' ij ∈[0,1],r' ij is the similarity coefficient; Step S14: Solve the transfer closure function t(R) of the fuzzy similarity coefficient matrix R and obtain t(R)=R n , and take different confidence levels to obtain different clustering results of driver behavior data; Step S15: storing the clustering result as a feature vector sample interval, continuing to collect new driver behavior data, and matching the new driver behavior data with the feature vector sample interval. If the match is successful, obtaining the classification result of the driver behavior data; if not, determining that the driver behavior data is irrelevant data; Step S16: Continue to collect the accumulated driver behavior data for the preset period, and merge the matching results of the new driver behavior data with the clustering results of the previously stored accumulated driver behavior data, and return to step S11 to repeat the step of performing fuzzy cluster analysis on the driver behavior data.

3. The vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm according to claim 1, characterized in that: The driver behavior data includes time consumption data for different turns at intersections and customary driving speeds for each speed-limited road section between each speed-limited node. The driver behavior data specifically includes: time consumption data for left turns, right turns, and straight-ahead driving at intersections during congested periods, time consumption data for left turns, right turns, and straight-ahead driving at intersections during non-congested periods, driving speeds between speed-limited nodes during congested and non-congested periods on holidays, and driving speeds between speed-limited nodes during congested and non-congested periods on non-holidays.

4. The vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm according to claim 1, characterized in that: The step S5 includes: when it is determined that the driver inputs the destination information and accepts the warning information, selecting the first battery swap navigation calculation mode; when it is determined that the driver inputs the destination information and does not accept the warning information, selecting the second battery swap navigation calculation mode; when it is determined that the driver does not input the destination information and accepts the warning information, selecting the third battery swap navigation calculation mode; when it is determined that the driver does not input the destination information and does not accept the warning information, selecting the fourth battery swap navigation calculation mode; wherein, When the first and third battery swap navigation calculation modes are selected, step S6 is directly executed. When the second and fourth battery swap navigation calculation modes are selected, the SOC value of the power battery in the working state is detected, and when the SOC value is less than the second preset threshold, step S6 is forcibly executed.

5. The vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm according to claim 4, characterized in that: The step of searching for a plurality of battery swap station distribution points according to the instantaneous vehicle information and the remaining cruising range in step S7 includes: When the first and second battery swap navigation calculation modes are selected, the input destination information is obtained, and the instantaneous location of the vehicle is used as the vertex of the sector. The center line of the sector is set along the direction of the destination. Under the condition of meeting the remaining cruising range, the sector coverage area of ​​0° to 180° is searched to obtain several battery swap station distribution points; When the third and fourth battery swap navigation calculation modes are selected, the vehicle's instantaneous location is used as the center, and the remaining cruising range condition is met. The circular coverage area of ​​0° to 360° is searched to obtain several battery swap station distribution points.

6. The vehicle intelligent battery swapping navigation method based on a multi-objective genetic algorithm according to claim 1, characterized in that: The average waiting time of the battery swap queue in step S7 is calculated using a multi-service station mixed queuing theory model, and the following formula is used to determine the average waiting time of the battery swap queue: Among them, W q is the average waiting time in the battery replacement queue, L q is the average length, λ e is the effective vehicle arrival rate.

7. A vehicle intelligent battery swapping navigation system based on a multi-objective genetic algorithm, characterized in that: include: A battery health detection module (10) is used to obtain the power battery loss value in real time; A driver behavior detection module (20) is used to collect and send driver behavior data to a database for fuzzy cluster analysis; A current detection module (30) is used to detect the initial SOC value of the power battery when the vehicle is started or paused using an open circuit voltage method, and to detect the SOC value of the power battery in a working state using an ampere-hour integration method; An early warning module (40) is used to send battery replacement early warning information; A satellite positioning module (50) is used to detect the instantaneous vehicle information of the vehicle in real time; The battery swap system data platform (60) is used to send data on the number of battery swap waiting points, the average waiting time for battery swapping, and the number of fully charged batteries at each battery swap station distribution point; A road supervision data platform (70) is used to send basic road data, road congestion data and path turning direction data of vehicles heading to various battery swap station distribution points; A wireless communication module (80) is used to transmit data on the number of vehicles waiting for battery swapping, the average waiting time in the battery swapping queue, the number of fully charged batteries, basic road data for vehicles traveling to various battery swapping stations, road congestion, and path turning direction data to a microcontroller (90); The microcontroller (90) is used to determine whether the power battery loss value is greater than the power battery loss preset value, and if so, control the warning module (40) to send a battery replacement warning message, and if not, control the current detection module (30) to detect the SOC value of the power battery and determine whether the SOC value is less than a first preset threshold value, and if so, control the warning module (40) to send a battery replacement warning message, and if not, return to control the battery health detection module (10) to obtain the power battery loss value in real time; and, after the warning module (40) sends the battery replacement warning message, determine whether the driver enters the destination information and Whether the driver accepts the warning information, and selects the corresponding battery swap navigation calculation mode according to the judgment result; after starting the selected battery swap navigation calculation mode, the microcontroller (90) is further used to calculate the remaining cruising range according to the SOC initial value and the SOC value, and search for a number of battery swap station distribution points according to the instantaneous vehicle information and the remaining cruising range, obtain the number of battery swap points in each battery swap station distribution point and the battery swap vehicle capacity of the battery swap station, establish a multi-service station mixed queuing theory model according to the number of battery swap points in the station and the battery swap vehicle capacity of the battery swap station, and calculate the average waiting time of the battery swap queue through the multi-service station mixed queuing theory model; It also obtains basic road data, road congestion, path turning direction data, and driver behavior data after fuzzy cluster analysis for vehicles traveling to each battery swap station distribution point; Based on the basic road data of the vehicle to each battery swap station distribution point, a travel time prediction model is established, and the vehicle travel time is predicted by the travel time prediction model; The driver's driving behavior habit category is determined based on the path turning direction data and the driver behavior data after fuzzy cluster analysis, and the total en route driving time of the vehicle is determined based on the driving behavior habit category and the vehicle's en route driving time, so as to establish a multi-objective optimization model based on the total en route driving time and the average waiting time in the battery swap queue. The multi-objective optimization model is a target optimization model including an objective function and constraints; wherein the objective function includes minimizing the total battery swap time of the user and maximizing driving comfort, wherein the total battery swap time of the user includes the total on-the-go driving time, the average waiting time in the battery swap queue and the battery swap operation time, and the battery swap operation time is a constant; driving comfort includes the number of turns on each path; the constraints include the vehicle's remaining power constraint, the road traffic congestion constraint, the length of the battery swap station queue constraint and the multi-step prediction constraint. The multi-objective optimization model is used to perform a multi-objective genetic optimization calculation to obtain the optimal battery swap station and the optimal path. The multi-objective optimization model is established, and the number of iterations t and the maximum number of iterations t are preset. max , and obtain the t-th generation parent population P t ; Get the fitness value, and adjust the parent population P according to the fitness value t The population chromosome is selected; the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Perform a crossover operation; perform a mutation operation on the parent population after the crossover to obtain the t-th generation offspring population Q t ; The t-th generation parent population P t and the t-th generation offspring population Q t Merge and use the non-dominated sorting method to obtain the t+1 generation parent population P t+1 ; Get the battery swap station and route allocation results; Determine whether the number of iterations is the maximum number of iterations t max If yes, then output the allocation result and get the optimal swap station and optimal path. If no, then return to the parent population P t The chromosomes with common gene sites in the population chromosomes form a pair and t Steps to perform crossover operation.

Images