New energy vehicle fleet management system and method based on stop point analysis

Through the new energy fleet management system based on stop point analysis, the configuration of stop point and charging piles is optimized, and the problem of inappropriate configurations of stop point and charging piles in the existing technology is solved, charging efficiency and fleet management efficiency are improved, and charging is ensured. Demand satisfaction and troubleshooting.

CN118863333BActive Publication Date: 2024-12-06TIANJIN HYDROGEN INVESTMENT NEW ENERGY VEHICLE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410819194.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-06-24
Publication Date
2024-12-06
Estimated Expiration
2044-06-24

AI Technical Summary

Technical Problem

The existing new energy fleet has unfit problems in the selection of stop points and the configuration of charging piles, which may cause the vehicle to be idle or unable to meet the needs when charging, affecting the charging efficiency and fleet management efficiency.

Method used

Through the new energy fleet management system based on stop point analysis, the multi-objective optimization model and charging demand prediction model are used to optimize the charging pile configuration of the stop point, and combined with genetic algorithms and machine learning algorithms, dynamically adjust the stop point position and the number of charging piles to monitor the Charge status and provide emergency maintenance solutions.

Benefits of technology

It improves the charging efficiency of the vehicle at the stop point, reduces the vehicle's stay pressure, ensures the satisfaction of charging needs, and promptly handles charging pile failures, improving the real-time and effectiveness of fleet management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118863333B_ABST
    Figure CN118863333B_ABST
Patent Text Reader

Abstract

The present invention discloses a new energy vehicle fleet management system and method based on stay point analysis, which relates to the technical field of new energy vehicle fleet management. A high-demand area is selected from a plurality of sub-areas according to the demand degree, and stay points are added in the high-demand area; the charging demand of vehicles in the stay points is predicted, and the corresponding charging demand is obtained, and the number of charging piles required for the stay points is determined according to the charging demand; a charging pile configuration scheme for the vehicle stay point is output using a multi-objective optimization model, different charging pile configuration schemes are tested, and efficiency coefficients are constructed from charging test data, and a plurality of charging pile configuration schemes are selected according to the efficiency coefficients; if a charging fault is determined according to monitoring data, a corresponding emergency maintenance scheme is output from a charging pile maintenance knowledge graph; when a plurality of vehicles enter the stay point for charging, the charging efficiency can be improved and the occupation of the stay point by the fleet vehicles can be reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of new energy vehicle fleet management, and in particular to a new energy vehicle fleet management system and method based on stop point analysis. Background Art

[0002] A new energy fleet is a fleet of new energy vehicles, which can include electric vehicles, hybrid vehicles and other types of new energy vehicles. The popularization and application of new energy fleets are of great significance in reducing pollution, reducing energy consumption, and promoting green and intelligent transportation. For example, in some logistics fleets, the operating cost per kilometer of using new energy heavy trucks is much lower than that of fuel vehicles, which makes new energy fleets attractive in terms of economic benefits.

[0003] In the Chinese invention patent with application publication number CN110633821A, an intelligent fleet management system is disclosed, including a control module, a positioning module connected to the control template, a battery management module, a vehicle condition monitoring module, a video processing module and a first wireless transmission module; and a fleet server, and a weather acquisition module, a road condition acquisition module, a second wireless transmission module and a terminal connected to the fleet server; the second wireless transmission module connects the maintenance station server, the charging station server and the first wireless transmission module. The beneficial effects of the present invention are real-time scheduling of vehicle operation, early warning planning of vehicle charging and maintenance, real-time monitoring of driver and vehicle conditions and establishment of database evaluation, improving the real-time and effectiveness of enterprise fleet management, reducing energy loss and increasing driving safety.

[0004] Combined with the above application and the contents of the prior art:

[0005] Usually, there are several new energy vehicles in a fleet, such as a taxi fleet or a bus fleet. If the new energy vehicle has a long driving route and consumes a lot of energy, it may need to be charged frequently. In this case, several parking points need to be set up in the activity area of ​​the fleet vehicles. The vehicles are charged after entering the parking points to ensure that the new energy vehicles can continue to drive. If there are few parking points in the vehicle activity area and the parking load is large, when managing the fleet vehicles, it is necessary to increase the parking points and configure charging piles at the parking points.

[0006] However, the existing stop point selection method mainly refers to the activity density of vehicles, and the number of charging piles is also set in proportion after determining the number of vehicles in the fleet. However, after determining the location of the stop point based on this method, there may be a situation where the vehicle stop load capacity and stop demand at the stop point are not matched. For example, the charging piles in the stop point may be idle or unable to meet the actual charging demand, resulting in vehicles having to queue up for charging.

[0007] To this end, the present invention provides a new energy vehicle fleet management system and method based on stop point analysis. Summary of the invention

[0008] 1. Technical issues to be resolved

[0009] In view of the deficiencies in the prior art, the present invention provides a new energy vehicle fleet management system and method based on stop point analysis, which adds stop points in high-demand areas; predicts the charging demand of vehicles in the stop points, obtains the corresponding charging demand, and determines the number of charging piles required for the stop points based on the charging demand; uses a multi-objective optimization model to output the charging pile configuration plan for the vehicle stop points, tests different charging pile configuration plans, and constructs an efficiency coefficient based on the charging test data, selects several charging pile configuration plans based on the efficiency coefficient, and if it is judged that there is a charging fault based on the monitoring data, the charging pile maintenance knowledge graph outputs the corresponding emergency maintenance plan; when several vehicles enter the stop point for charging, the charging efficiency can be improved, thereby solving the technical problems raised in the background technology.

[0010] (II) Technical solution

[0011] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0012] A new energy vehicle fleet management method based on stop point analysis includes dividing the high-frequency stop area of ​​the vehicle fleet into several sub-areas, generating a stop load degree Hop from the traffic status data in the sub-areas, and if the stop load degree Hop exceeds a load threshold, issuing a stop point selection instruction to the outside;

[0013] The parking and charging data of the fleet vehicles and related data are collected, and a vehicle charging data set is constructed after preprocessing. The demand Xot is constructed from the related data in the sub-area. According to the demand Xot, high-demand areas are selected from several sub-areas, and parking points are added in the high-demand areas. The generation method of the demand Xot is as follows: the passing frequency Po and the terrain flatness Do are linearly normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following formula:

[0014]

[0015] Weight coefficient: 0≤χ≤1, 0≤ρ≤1, and χ+ρ=1;

[0016] After selecting the key factors that affect the charging efficiency of charging piles, use the trained charging demand prediction model to predict the charging demand of vehicles at the stop point, obtain the corresponding charging demand, and determine the number of charging piles required at the stop point based on the charging demand;

[0017] Use a multi-objective optimization model to output the charging pile configuration plan for the vehicle stop point, use the vehicle charging digital twin model to test different charging pile configuration plans, and construct the efficiency coefficient Vot from the charging test data. Select several charging pile configuration plans based on the efficiency coefficient Vot;

[0018] When fleet vehicles are charging at a stop, the charging status is monitored. If a charging fault is detected based on the monitoring data, the charging pile maintenance knowledge graph will output a corresponding emergency maintenance plan after extracting the fault characteristics, and issue an alarm command to the outside.

[0019] Furthermore, the activity area of ​​the target vehicle in the fleet when performing the transfer task is obtained, and an electronic map covering the activity area is constructed. The driving data and historical location data of the target vehicle in the fleet are recorded in the activity area, and the current stop position of the target vehicle is queried and obtained. The stop position is marked on the electronic map and a heat map of the target vehicle's stop is generated, and the high-frequency stop areas of the target vehicle are identified from the heat map.

[0020] Furthermore, the density of the stay points in each sub-area is obtained, and the traffic flow data on the target vehicle driving path in each sub-area is queried and obtained. The obtained stay point density and traffic flow data are summarized to construct a traffic status data set; the stay load Hop is generated from the data in the traffic status data set, wherein the stay point density Ot and the traffic volume Pt ​​are linearly normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following formula:

[0021]

[0022] Among them, k 2 , k 1 is the weight coefficient, 0≤k 1 ≤1,0≤k 2 ≤1, and k 1 +k 2 =1,Pt i is the traffic volume in the ith sub-area, Pt a is the mean traffic volume; Ot i is the density of stay points in the ith sub-area Ot, Ot a is the mean density of the dwell points.

[0023] Furthermore, after receiving the stop point selection instruction, data is collected in the activity area, including vehicle data, charging data, and charging pile data; the collected data is preprocessed, and the preprocessed data is aggregated to construct a vehicle charging data set; the demand Xot for adding stop points is constructed based on the vehicle's driving data in the sub-area, and if the demand Xot in the sub-area exceeds the demand threshold, the corresponding sub-area is determined as a high-demand area.

[0024] Furthermore, several backup points are selected in the high-demand area, and a multi-objective optimization model is obtained by using genetic algorithm training. Maximizing the utilization rate of charging piles and minimizing the waiting time of vehicles are used as optimization goals. The vehicle capacity in the backup points and the number of installable charging piles are used as constraints, and new stop points are selected from several backup points.

[0025] Furthermore, when the fleet vehicles are in the charging state, the charging status data of the vehicles are collected, and the charging efficiency is used as the dependent variable, and the charging conditions when the vehicles are charging are used as the independent variables. The influence of each charging condition on the charging efficiency is obtained through multiple linear regression analysis, and the key factors of charging are screened out within several charging conditions according to the influence degree, and the key factors of the data are output.

[0026] Furthermore, after obtaining data of different specifications of charging piles, a multi-objective optimization model is obtained by using genetic algorithm training, with maximum charging pile utilization and minimum vehicle waiting time as optimization goals, and the number of charging piles, power supply capacity and land area of ​​the stopover point as constraints, to optimize the configuration of the charging piles and obtain the current charging pile configuration plan;

[0027] Use machine learning algorithm training to obtain the vehicle charging digital twin model, and take the charging pile configuration scheme as input. The vehicle charging digital twin model simulates the charging effects of different charging pile configuration schemes and obtains the corresponding test data. The obtained test data is summarized to construct a charging test data set.

[0028] Furthermore, the efficiency coefficient Vot is constructed from the test data in the charging test data set, wherein the waiting time C and the charging efficiency P are normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following method:

[0029]

[0030] Where i is the serial number of the vehicle that stops during the test period, i = 1, 2, ..., k; weight coefficient: 0 ≤ χ ≤ 1, 0 ≤ ρ ≤ 1, and χ + ρ = 1, C i The waiting time for charging the i-th parked vehicle, C a is the mean of k waiting times, P iis the charging efficiency of the i-th parked vehicle, P a is the mean of k charging efficiencies;

[0031] Several charging pile configuration schemes are ranked according to the efficiency coefficient Vot, and the one with the highest ranking is taken as the target scheme. The charging piles are installed at the stop point according to the target scheme.

[0032] Furthermore, after the installation of the charging pile is completed, the charging status of the vehicle is monitored and corresponding monitoring data is obtained, the monitoring data is identified, and it is determined whether there is an abnormality during the charging operation. If there is an abnormality, a fault diagnosis instruction is issued;

[0033] Build an initial model based on the machine learning algorithm, use historical fault data to train the initial model, obtain the fault diagnosis model, and after receiving the fault diagnosis instruction, use the fault diagnosis model to diagnose the charging pile. If the charging pile has an operating fault, extract the corresponding fault characteristics; use charging pile emergency maintenance as the target word, and pre-build the charging pile maintenance knowledge graph.

[0034] The new energy fleet management system based on the stop point analysis includes: an area screening unit, which divides the high-frequency stop area of ​​the fleet vehicles into several sub-areas, generates the stop load degree Hop from the traffic status data in the sub-areas, and sends a stop point selection instruction to the outside if the stop load degree Ho exceeds the load threshold;

[0035] The demand analysis unit collects the parking and charging data and related data of the fleet vehicles, constructs a vehicle charging data set after preprocessing, and constructs the demand Xot from the related data in the sub-area. According to the demand Xot, high-demand areas are selected in several sub-areas, and parking points are added in the high-demand areas;

[0036] The demand prediction unit selects the key factors that affect the charging efficiency of the charging piles, uses the trained charging demand prediction model to predict the charging demand of the vehicles at the stop point, obtains the corresponding charging demand, and determines the number of charging piles required at the stop point based on the charging demand;

[0037] The scheme screening unit uses a multi-objective optimization model to output the charging pile configuration scheme for the vehicle stop point, uses the vehicle charging digital twin model to test different charging pile configuration schemes, and constructs the efficiency coefficient Vot based on the charging test data, and selects several charging pile configuration schemes based on the efficiency coefficient Vot;

[0038] The fault alarm unit monitors the charging status when the fleet vehicles are charging at the stop point. If a charging fault is determined based on the monitoring data, the charging pile maintenance knowledge graph will output the corresponding emergency maintenance plan after extracting the fault characteristics, and issue an alarm command to the outside.

[0039] (III) Beneficial effects

[0040] The present invention provides a new energy fleet management system and method based on stop point analysis, which has the following beneficial effects:

[0041] 1. Construct the dwell load degree Hop of the target vehicle when it stops, and judge whether the dwell pressure of the fleet vehicles in the high-frequency dwelling area is large according to the dwell load degree Ho. If the dwell pressure is large, reselect a new dwelling point of the target vehicle, so as to reduce the dwell pressure of the vehicle during the dwell management of the fleet vehicles.

[0042] 2. When it is necessary to set up a new stop point in a sub-area, a high-demand area is selected from several sub-areas. When a new stop point is selected in a high-demand area, it is more targeted and reliable.

[0043] 3. The position of the backup point is adjusted and selected through the genetic algorithm, and the selected backup point is used as the new stop point, so that the comprehensive effect of the new stop point is better, and the pressure of the vehicle staying in the high-frequency stop area is smaller.

[0044] 4. Through regression analysis, we screen several environmental conditions and select key factors from them. When it is necessary to improve the charging efficiency, we can implement targeted processing. The charging demand prediction model obtained through training can predict the charging demand at the stop point. Therefore, when the corresponding charging demand is obtained, we can roughly determine the number of charging piles to be installed at the stop point.

[0045] 5. The vehicle charging digital twin model simulates and tests several charging configuration schemes, and constructs the efficiency coefficient Vot based on the acquired test data. The acquired charging pile configuration scheme is selected based on the efficiency coefficient Vot. When several vehicles enter the stop point for charging, the charging efficiency can be improved and the occupancy of the stop point by the fleet vehicles can be reduced.

[0046] 6. After the vehicle enters the charging state, the charging state is monitored and the corresponding monitoring data is obtained. The trained fault diagnosis model is used to diagnose the charging pile to determine whether there is a fault. If there is a fault, it can be handled in time after the type of fault is determined;

[0047] 7. The corresponding emergency maintenance plan is output from the charging pile maintenance knowledge graph. When the charging pile needs to be processed, the emergency maintenance plan is used as a reference. When there may be charging problems with the charging pile in the stop point, it can be processed in time, and the charging efficiency of the vehicle is improved when the vehicle is parked at the stop point. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a schematic flow chart of a new energy vehicle fleet management method based on stop point analysis of the present invention;

[0049] Figure 2 It is a structural schematic diagram of a new energy vehicle fleet management system based on stop point analysis of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] See also Figure 1 The present invention provides a new energy vehicle fleet management method based on stay point analysis, comprising:

[0052] Step 1: Divide the high-frequency stop area of ​​the fleet vehicles into several sub-areas, generate the stop load degree Hop based on the traffic status data in the sub-areas, and if the stop load degree Hop exceeds the load threshold, issue a stop point selection instruction to the outside;

[0053] The step 1 includes the following contents:

[0054] Step 101: obtaining the activity area of ​​the target vehicle in the fleet when performing the transfer task, and constructing an electronic map covering the activity area, recording the driving data and historical location data of the target vehicle in the fleet in the activity area, and querying to obtain the current stop position of the target vehicle, marking the stop position on the electronic map and generating a heat map of the target vehicle's stop, and identifying the high-frequency stop area of ​​the target vehicle from the heat map;

[0055] When in use, after obtaining the driving data of the vehicles in the fleet, the vehicle position of the target vehicle is determined by the driving data and the corresponding stop data, and the high-frequency stop area is selected in the activity area according to the stop position of the vehicle, so that targeted processing can be carried out when the stop and charging management of the fleet vehicles is carried out;

[0056] Step 102: divide the high-frequency stay area into several sub-areas, mark each sub-area, obtain the stay point density in each sub-area, query and obtain the traffic flow data on the target vehicle driving path in each sub-area, summarize the obtained stay point density and traffic flow data, and construct a traffic status data set;

[0057] Step 103: Generate the dwell load Hop from the data in the traffic status data set, wherein the dwell point density Ot and the traffic volume Pt ​​are linearly normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following formula:

[0058]

[0059] Among them, k 2 , k 1 is the weight coefficient, 0≤k 1 ≤1,0≤k 2 ≤1, and k 1 +k 2 =1, which can be obtained by referring to the hierarchical analysis method; Pt i is the traffic volume in the ith sub-area, Pt a is the mean traffic volume; Ot i is the density of stay points in the ith sub-area Ot, Ot a is the mean density of the stay points;

[0060] Based on historical data and management expectations for fleet vehicle stops, a load threshold is pre-set; if the stop load Hop exceeds the preset load threshold, it means that the target vehicle stop load in the high-frequency stop area is large, and a new stop point needs to be selected near the high-frequency stop area and its adjacent areas, and a stop point selection instruction is issued;

[0061] The Analytic Hierarchy Process (AHP) is a decision-making method that decomposes decision-related elements into levels such as goals, criteria, and plans, and conducts qualitative and quantitative analysis on this basis. It is particularly suitable for dealing with target systems with hierarchical and staggered evaluation indicators, and when the target value is difficult to describe quantitatively, the AHP is an effective decision-making tool.

[0062] The core of the hierarchical analysis method is to decompose the decision problem into multiple levels to form a hierarchical structure, which usually includes the target level, the criterion level, the sub-criterion level and the solution level. By solving the eigenvector of the judgment matrix, the priority weight of each element of each level to an element of the previous level is obtained, and the final weight of each alternative solution to the total goal is recursively merged by the weighted sum method, so as to find the optimal solution.

[0063] When using, combine the contents in steps 101 to 103:

[0064] After the high-frequency stop area has been selected, it is divided into several sub-areas. According to the stop point location in the sub-area and the corresponding traffic flow data, the stop load degree Hop of the target vehicle when it stops is constructed. According to the stop load degree Ho, it is judged whether the stop pressure of the fleet vehicles in the high-frequency stop area is large. If the stop pressure is large, it is necessary to reselect a new stop point for the target vehicle, so as to reduce the stop pressure of the vehicle when the fleet vehicles are managed.

[0065] However, the existing stop point selection method mainly refers to the activity density of vehicles, and the number of charging piles is also set in proportion after determining the number of vehicles in the fleet. However, after determining the location of the stop point based on this method, the vehicle stop load and stop demand at the stop point may not match. For example, the charging piles in the stop point may be idle or unable to meet the actual charging demand, resulting in vehicles having to queue up for charging.

[0066] Step 2: Collect the parking and charging data and related data of the fleet vehicles, construct a vehicle charging data set after preprocessing, and construct the demand Xot from the related data in the sub-area. According to the demand Xot, select high-demand areas in several sub-areas and add parking points in the high-demand areas;

[0067] The step 2 includes the following contents:

[0068] Step 201: After receiving the stop point selection instruction, data collection is performed in the activity area, for example, vehicle data, including location data, driving speed, driving mileage, acceleration, power percentage, battery voltage, current, and charging temperature; charging data, including charging status, charging time, and charging power; charging pile data, including charging pile usage status, real-time charging power, and power of each charge;

[0069] Preprocess the collected data. The preprocessing methods include: data cleaning: processing missing values, processing abnormal values, processing duplicate values, correcting erroneous values: correcting or replacing obviously erroneous data. Data conversion, data integration, data dimension reduction, data smoothing, etc.; summarize the preprocessed data to construct a vehicle charging data set;

[0070] During use, when it is necessary to continue selecting a stop point, the reliability of the stop point selection can be increased after data collection.

[0071] Step 202: construct the demand Xot of adding a stop point based on the driving data of the vehicle in the sub-area, and select a high-demand area from several sub-areas based on the demand Xot, wherein the vehicle passing frequency Po and the terrain flatness Do in the sub-area are linearly normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following formula:

[0072]

[0073] Weight coefficient: 0≤χ≤1, 0≤ρ≤1, and χ+ρ=1; the weight coefficient can be obtained by referring to the hierarchical analysis method;

[0074] According to historical data and the management requirements for vehicle parking demand, a demand threshold is pre-set; if the demand Xot in a sub-area exceeds the demand threshold, it means that the parking demand in the corresponding sub-area is large, and the corresponding sub-area is determined as a high-demand area;

[0075] During use, when a new stay point needs to be set in a sub-area, if the number of stay points to be set is limited or the locations are limited, a high-demand area is selected from several sub-areas, and a new stay point is selected in the high-demand area, which is more targeted and reliable.

[0076] Step 203: select several spare points in the high-demand area, use genetic algorithm training to obtain a multi-objective optimization model, take maximizing the utilization rate of charging piles and minimizing the waiting time of vehicles as optimization goals, take the vehicle capacity in the spare points and the number of charging piles that can be installed as constraints, and select a new stop point from the several spare points;

[0077] When using, combine the contents in steps 201 and 202:

[0078] After screening out high-risk areas and the backup points within them, the locations of the backup points are adjusted and selected through a genetic algorithm based on the collected data, and the selected backup points are used as new stop points to make the comprehensive effect of the new stop points better. When vehicles stop in high-frequency stop areas, the pressure of stopping is smaller and the degree of matching is higher.

[0079] Step 3: After selecting the key factors that affect the charging efficiency of the charging pile, use the trained charging demand prediction model to predict the charging demand of the vehicles at the stop point, obtain the corresponding charging demand, and determine the number of charging piles required at the stop point based on the charging demand;

[0080] The step three includes the following contents:

[0081] Step 301: When the fleet vehicle is in a charging state, collect the vehicle's charging state data, including charging efficiency, charging time, etc., as well as the weather and ambient temperature during charging, etc.; take the charging efficiency as the dependent variable, take the charging condition during vehicle charging as the independent variable, obtain the influence of each charging condition on the charging efficiency through multiple linear regression analysis, select the key factors during charging from a number of charging conditions according to the influence, and output the key factors of the data;

[0082] Step 302: construct an initial model using a convolutional neural network, train the initial model using sample data, obtain a charging demand prediction model, use the trained charging demand prediction model to predict the charging demand of vehicles at the stop point, and obtain the corresponding charging demand, and determine the number of charging piles required for each stop point based on the charging demand, combined with factors such as charging speed, power and usage rate;

[0083] When using, combine the contents in steps 301 and 302:

[0084] When managing vehicles, after collecting various charging data, regression analysis is used to screen them within several environmental conditions, and key factors are selected from them. When it is necessary to improve the charging efficiency, targeted processing can be achieved. Furthermore, when it is necessary to consider installing charging piles at the stop points, the charging demand prediction model obtained through training is used to predict the charging demand at the stop points. Therefore, when the corresponding charging demand is obtained, the number of charging piles to be installed at the stop points can be roughly determined.

[0085] Step 4: Use the multi-objective optimization model to output the charging pile configuration plan for the vehicle stop point, use the vehicle charging digital twin model to test different charging pile configuration plans, and construct the efficiency coefficient Vot from the charging test data, and select several charging pile configuration plans based on the efficiency coefficient Vot;

[0086] The step 4 includes the following contents:

[0087] Step 401: After obtaining data of different specifications of charging piles, a multi-objective optimization model is obtained by using genetic algorithm training, with maximum charging pile utilization and minimum vehicle waiting time as optimization goals, and the number of charging piles, power supply capacity and land area of ​​the stopover point as constraints, to optimize the configuration of the charging piles and obtain the current charging pile configuration plan;

[0088] When in use, if it is necessary to configure the charging piles at the stop point, the genetic algorithm is used to optimize the charging configuration plan to obtain the optimized charging pile configuration plan. When the charging piles are installed, the installation of the charging piles is more reasonable.

[0089] Step 402: Use a machine learning algorithm to train and obtain a vehicle charging digital twin model. Take the configuration scheme of the charging pile as input. The vehicle charging digital twin model simulates the charging effects of different charging pile configuration schemes and obtains corresponding test data, including: collecting charging pile utilization, waiting time and charging efficiency when the vehicle is charging; and aggregating the obtained test data to construct a charging test data set.

[0090] Step 403: construct the efficiency coefficient Vot from the test data in the charging test data set, wherein the waiting time C and the charging efficiency P are normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following method:

[0091]

[0092] Where i is the serial number of the vehicle that stops during the test period, i = 1, 2, ..., k; weight coefficient: 0 ≤ χ ≤ 1, 0 ≤ ρ ≤ 1, and χ + ρ = 1, and its value is consistent with the previous value; C i The waiting time for charging the i-th parked vehicle, C a is the mean of k waiting times, P i is the charging efficiency of the i-th parked vehicle, P a is the mean of k charging efficiencies;

[0093] Sort several charging pile configuration schemes by efficiency coefficient Vot, take the top one as the target scheme, and install the charging piles at the stop point according to the target scheme;

[0094] When used, combine the contents in steps 401 to 403:

[0095] After obtaining several charging pile configuration schemes, the vehicle charging digital twin model simulates and tests the several charging configuration schemes, and constructs the efficiency coefficient Vot based on the obtained test data. The obtained charging pile configuration schemes are selected based on the efficiency coefficient Vot, and the best configuration scheme is selected. Therefore, when several vehicles enter the stop point and charge at the stop point, the charging efficiency can be improved and the occupancy of the stop point by the fleet vehicles can be reduced.

[0096] Step 5: When the fleet vehicles are charging at the stop point, the charging status is monitored. If a charging fault is detected based on the monitoring data, the charging pile maintenance knowledge graph will output the corresponding emergency maintenance plan after extracting the fault characteristics, and issue an alarm command to the outside.

[0097] The step five includes the following contents:

[0098] Step 501: After the installation of the charging pile is completed, the charging status of the vehicle is monitored and corresponding monitoring data is obtained; after setting abnormal standards for various parameters, the monitoring data is identified to determine whether there is an abnormality during the charging operation, and if there is an abnormality, a fault diagnosis instruction is issued;

[0099] An initial model is constructed based on a machine learning algorithm, and the historical fault data is used to train the initial model to obtain a fault diagnosis model. After receiving a fault diagnosis instruction, the fault diagnosis model is used to perform fault diagnosis on the charging pile. If there is an operating fault in the charging pile, the corresponding fault features are extracted.

[0100] During use, after the configuration and installation of the charging pile is completed and the vehicle enters the charging state, the charging state is monitored and the corresponding monitoring data is obtained. The trained fault diagnosis model is used to diagnose the charging pile to determine whether there is a fault. If there is a fault, it can be handled in time after the type of fault is determined;

[0101] Step 502: Taking charging pile emergency maintenance as the target word, after deep retrieval and entity relationship building, a charging pile maintenance knowledge graph is pre-built. Using the trained matching model, according to the correspondence between the emergency maintenance plan and the fault characteristics, the charging pile maintenance knowledge graph outputs the corresponding emergency maintenance plan, wherein the emergency maintenance plan includes: troubleshooting, equipment replacement, backup plan, etc. At this time, an alarm instruction is issued to the outside;

[0102] When using, combine the contents in steps 501 and 502:

[0103] When it is necessary to maintain the charging pile at the stop point, the corresponding emergency maintenance plan is output from the charging pile maintenance knowledge graph through fault feature extraction. When the charging pile needs to be processed, the corresponding emergency maintenance plan can be quickly obtained. With the emergency maintenance plan as a reference, when there may be charging problems with the charging pile at the stop point, it can be processed in time, thereby improving the charging efficiency of the vehicle when the vehicle is parked at the stop point.

[0104] The construction method of the charging pile maintenance knowledge graph mainly includes the following steps:

[0105] Data collection and preprocessing: Data collection: Obtain data related to charging pile maintenance from various data sources, such as web text, databases, APIs, etc. This can be achieved through crawlers, API calls, or database exports.

[0106] Data preprocessing: Clean, transform and integrate the collected data to remove noise, stop words and punctuation marks, and reduce the impact of irrelevant information. In addition, operations such as word segmentation, part-of-speech tagging, named entity recognition, and attribute and relationship extraction are required to convert the data into the basic units in the knowledge graph - entities, attributes and relationships.

[0107] Entity recognition and linking: Entity recognition: Map the named entities in the text, such as charging piles, fault types, maintenance tools, etc., to entity nodes in the knowledge graph. This can be achieved through methods such as rule engines, machine learning or deep learning.

[0108] Entity linking: Find nodes representing the same entity in different data sources and connect them. This can be achieved by methods such as name similarity, attribute similarity, or relationship linking.

[0109] Attribute and relationship extraction: Attribute extraction: Extract attribute information related to entities from text, such as the type, power, manufacturer, etc. of charging piles. This can also be achieved through rule engines, machine learning, or deep learning.

[0110] Relationship extraction: Extract relationship information between entities from text, such as the relationship between charging piles and fault types, the relationship between maintenance tools and fault solutions, etc.

[0111] Knowledge graph storage and query: Storage: Choose an appropriate way to store the constructed knowledge graph for subsequent query and use. Common storage methods include using RDF (Resource Description Framework) and other standardized storage formats for storage.

[0112] Query: Provides a convenient query interface, allowing users to retrieve relevant knowledge on charging pile maintenance as needed.

[0113] See also Figure 2 The present invention provides a new energy vehicle fleet management system based on stay point analysis, including:

[0114] The area screening unit divides the high-frequency stop area of ​​the fleet vehicles into several sub-areas, and generates the stop load degree Hop based on the traffic status data in the sub-areas. If the stop load degree Hop exceeds the load threshold, a stop point selection instruction is issued to the outside.

[0115] The demand analysis unit collects the parking and charging data and related data of the fleet vehicles, constructs a vehicle charging data set after preprocessing, and constructs the demand Xot from the related data in the sub-area. According to the demand Xot, high-demand areas are selected in several sub-areas, and parking points are added in the high-demand areas;

[0116] The demand prediction unit selects the key factors that affect the charging efficiency of the charging piles, uses the trained charging demand prediction model to predict the charging demand of the vehicles at the stop point, obtains the corresponding charging demand, and determines the number of charging piles required at the stop point based on the charging demand;

[0117] The scheme screening unit uses a multi-objective optimization model to output the charging pile configuration scheme for the vehicle stop point, uses the vehicle charging digital twin model to test different charging pile configuration schemes, and constructs the efficiency coefficient Vot based on the charging test data, and selects several charging pile configuration schemes based on the efficiency coefficient Vot;

[0118] The fault alarm unit monitors the charging status when the fleet vehicles are charging at the stop point. If a charging fault is determined based on the monitoring data, the charging pile maintenance knowledge graph will output the corresponding emergency maintenance plan after extracting the fault characteristics, and issue an alarm command to the outside.

[0119] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0120] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0121] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0122] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0123] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0124] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0125] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0126] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. New energy vehicle fleet management method based on stop point analysis, Features: include, The high-frequency stop area of ​​the fleet vehicles is divided into several sub-areas, and the stop load degree Hop is generated from the traffic status data in the sub-areas. If the stop load degree Hop exceeds the load threshold, a stop point selection instruction is issued to the outside. The parking and charging data of the fleet vehicles and related data are collected, and a vehicle charging data set is constructed after preprocessing. The demand Xot is constructed from the related data in the sub-area. According to the demand Xot, high-demand areas are selected from several sub-areas, and parking points are added in the high-demand areas. The generation method of the demand Xot is as follows: the passing frequency Po and the terrain flatness Do are linearly normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following formula: Weight coefficient: 0≤χ≤1, 0≤ρ≤1, and χ+ρ=1; After selecting the key factors that affect the charging efficiency of charging piles, use the trained charging demand prediction model to predict the charging demand of vehicles at the stop point, obtain the corresponding charging demand, and determine the number of charging piles required at the stop point based on the charging demand; Use a multi-objective optimization model to output the charging pile configuration plan for the vehicle stop point, use the vehicle charging digital twin model to test different charging pile configuration plans, and construct the efficiency coefficient Vot from the charging test data. Select several charging pile configuration plans based on the efficiency coefficient Vot; When fleet vehicles are charging at a stop, the charging status is monitored. If a charging fault is detected based on the monitoring data, the charging pile maintenance knowledge graph will output a corresponding emergency maintenance plan after extracting the fault characteristics, and issue an alarm command to the outside.

2. The new energy vehicle fleet management method based on stay point analysis according to claim 1, Features: Obtain the activity area of ​​the target vehicle in the fleet when performing the transfer task, and build an electronic map covering the activity area. Record the driving data and historical location data of the target vehicle in the fleet in the activity area, query and obtain the current stop position of the target vehicle, mark the stop position on the electronic map and generate a heat map of the target vehicle's stop, and identify the high-frequency stop areas of the target vehicle from the heat map.

3. The new energy vehicle fleet management method based on stay point analysis according to claim 2, Features: Obtain the density of the stay points in each sub-area, and query and obtain the traffic flow data on the target vehicle's driving path in each sub-area, summarize the obtained stay point density and traffic flow data, and construct a traffic status data set; The dwell load Hop is constructed from the traffic status data set, where the dwell point density Ot and the traffic volume Pt ​​are linearly normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following formula: Among them, k 2 , k 1 is the weight coefficient, 0≤k 1 ≤1,0≤k 2 ≤1, and k 1 +k 2 =1,Pt i is the traffic volume in the ith sub-area, Pt a is the mean traffic volume; Ot i is the density of stay points in the ith sub-area Ot, Ot a is the mean density of the dwell points.

4. The new energy vehicle fleet management method based on stay point analysis according to claim 3, Features: After receiving the stop point selection instruction, data is collected in the activity area, including at least vehicle data, charging data and charging pile data; the collected data is preprocessed, and the preprocessed data is aggregated to construct a vehicle charging data set; The demand Xot of the added stop point is constructed based on the driving data of the vehicle in the sub-area. If the demand Xot in the sub-area exceeds the demand threshold, the corresponding sub-area is determined as a high-demand area.

5. The new energy vehicle fleet management method based on stay point analysis according to claim 4, Features: Several backup points are selected in high-demand areas, and a multi-objective optimization model is obtained by genetic algorithm training. The optimization goals are to maximize the utilization of charging piles and minimize the waiting time of vehicles. The vehicle capacity in the backup points and the number of charging piles that can be installed are used as constraints to select new stop points from several backup points.

6. The new energy vehicle fleet management method based on stay point analysis according to claim 1, Features: When the fleet vehicles are in the charging state, the charging status data of the vehicles is collected, with the charging efficiency as the dependent variable and the charging conditions when the vehicles are charging as the independent variables. The influence of each charging condition on the charging efficiency is obtained through multiple linear regression analysis. The key factors of charging are screened out within several charging conditions based on the influence, and the key factors of the data are output.

7. The new energy vehicle fleet management method based on stay point analysis according to claim 6, Features: After obtaining data on different specifications of charging piles, a genetic algorithm is used to train a multi-objective optimization model. The maximum charging pile utilization rate and the minimum vehicle waiting time are used as optimization goals. The number of charging piles, power supply capacity, and land area of ​​the stopover point are used as constraints to optimize the configuration of charging piles and obtain the current charging pile configuration plan. Use machine learning algorithm training to obtain the vehicle charging digital twin model, and take the charging pile configuration scheme as input. The vehicle charging digital twin model simulates the charging effects of different charging pile configuration schemes and obtains the corresponding test data. The obtained test data is summarized to construct a charging test data set.

8. The new energy vehicle fleet management method based on stay point analysis according to claim 7, Features: The efficiency coefficient Vot is constructed from the test data in the charging test data set, where the waiting time C and the charging efficiency P are normalized, and the corresponding data values ​​are mapped to the interval [0,1], and then according to the following method: Where i is the serial number of the vehicle that stops during the test period, i = 1, 2, ..., k; weight coefficient: 0 ≤ χ ≤ 1, 0 ≤ ρ ≤ 1, and χ + ρ = 1, C i The waiting time for charging the i-th parked vehicle, C a is the mean of k waiting times, P i is the charging efficiency of the i-th parked vehicle, P a is the mean of k charging efficiencies; Several charging pile configuration schemes are ranked according to the efficiency coefficient Vot, and the one with the highest ranking is taken as the target scheme. The charging piles are installed at the stop point according to the target scheme.

9. The new energy vehicle fleet management method based on stay point analysis according to claim 8, Features: After the installation of the charging pile is completed, the charging status of the vehicle is monitored and the corresponding monitoring data is obtained. The monitoring data is identified to determine whether there is any abnormality during the charging operation. If there is any abnormality, a fault diagnosis instruction is issued; Build a fault diagnosis model based on machine learning algorithms. After receiving the fault diagnosis instruction, use the fault diagnosis model to diagnose the charging pile. If there is an operating fault in the charging pile, extract the corresponding fault characteristics. Taking charging pile emergency maintenance as the target word, a charging pile maintenance knowledge graph is pre-constructed.

10. New energy fleet management system based on stop point analysis, Features: include: The area screening unit divides the high-frequency stop area of ​​the fleet vehicles into several sub-areas, and generates the stop load degree Hop based on the traffic status data in the sub-areas. If the stop load degree Hop exceeds the load threshold, a stop point selection instruction is issued to the outside. The demand analysis unit collects the parking and charging data and related data of the fleet vehicles, constructs a vehicle charging data set after preprocessing, and constructs the demand Xot from the related data in the sub-area. According to the demand Xot, high-demand areas are selected in several sub-areas, and parking points are added in the high-demand areas; The demand prediction unit selects the key factors that affect the charging efficiency of the charging piles, uses the trained charging demand prediction model to predict the charging demand of the vehicles at the stop point, obtains the corresponding charging demand, and determines the number of charging piles required at the stop point based on the charging demand; The scheme screening unit uses a multi-objective optimization model to output the charging pile configuration scheme for the vehicle stop point, uses the vehicle charging digital twin model to test different charging pile configuration schemes, and constructs the efficiency coefficient Vot based on the charging test data, and selects several charging pile configuration schemes based on the efficiency coefficient Vot; The fault alarm unit monitors the charging status when the fleet vehicles are charging at the stop point. If a charging fault is determined based on the monitoring data, the charging pile maintenance knowledge graph will output the corresponding emergency maintenance plan after extracting the fault characteristics, and issue an alarm command to the outside.