Electric vehicle charging path planning method and device

Abstract

The present invention discloses a method and device for planning a charging route for electric vehicles. By integrating vehicle status, user habits and preferences, and historical trajectory data, a constant current and constant voltage segmented charging model is constructed to accurately calculate the charging time cost, and a hierarchical analysis method is used to quantify the user's charging preference weight. A set of accessible charging routes is generated based on the remaining mileage, and the routes are scored and ranked using the TOPSIS multi-criteria decision model to achieve personalized recommendations. The device integrates BMS power monitoring, CAN bus communication, and cloud computing modules to support real-time road condition feedback and personalized route push. The system dynamically corrects charging curve parameters and preference models by recording user selection behavior, forming a closed-loop optimization mechanism. Compared with traditional solutions, the present invention takes into account both charging efficiency and user convenience, reduces decision-making complexity, and is suitable for intelligent charging decision support in dynamic traffic environments.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging navigation, and in particular to a method and device for planning an electric vehicle charging path. Background Art

[0002] In the past three years in my country, the number of electric vehicle consumers has continued to rise, and the number of electric vehicles in use has steadily increased. The increase in electric vehicles has led to long charging queues, and because some users do not have accurate information about the battery level of their vehicles, they often run out of battery power while driving, which increases users' "range anxiety" during driving and seriously affects their experience. Moreover, the current "range anxiety" is mainly reflected in time. People hope to meet their needs for travel and daily commuting in terms of time, and relatively speaking, their concern for the economic cost of charging has decreased. A reasonable electric vehicle charging path planning auxiliary method and device can provide users with a more personalized charging plan based on their usage habits, reduce their "range anxiety", and at the same time reduce their time cost, making a huge contribution to improving their experience.

[0003] Existing charging planning solutions do not use fitting curves to predict charging time, do not integrate user charging habits and preferences, cannot accurately calculate charging time costs, and lack personalized charging planning for users.

[0004] This paper constructs a dynamic time prediction mechanism based on a constant current-constant voltage segmented charging model. By fitting the battery charging curve, it accurately calculates the charging time for different power intervals. It integrates user historical trajectory and charging behavior data, and uses the analytic hierarchy process to establish a dynamic preference weight model. This quantifies implicit needs such as idle time utilization and charging station usage habits. The TOPSIS method combines multiple objective parameters such as real-time traffic conditions, charging curve delay, and user preferences to generate a hierarchical recommendation scheme that balances time efficiency and personal convenience. This reduces the user's decision-making burden while improving the rationality of route planning, effectively balancing charging efficiency, time cost, and user convenience needs. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the purpose of the present invention is to provide an electric vehicle charging route planning method, which can combine the user's charging habits, make reasonable inferences about the user's driving behavior and future behavior, actively push charging assistance plans, and provide users with corresponding route planning, thereby reducing users' mileage anxiety when using electric vehicles.

[0006] In order to achieve the above objectives, the present invention proposes the following solutions:

[0007] A method for planning an electric vehicle charging route includes a real-time power monitoring module based on a single-chip microcomputer, an information aggregation module connected to all vehicle modules via a CAN bus, an on-board ECU module based on the AUTOSAR protocol and the RH850 chip, a human-computer interaction module that supports the active push of charging assistance plans, and a cloud server that supports information storage and algorithm analysis.

[0008] The single chip microcomputer-based real-time power monitoring module supports being embedded in the vehicle's BMS to obtain the vehicle's current real-time status and battery charging characteristic data.

[0009] The information aggregation module, which connects all vehicle modules via the CAN bus, acquires real-time vehicle information, reports it to the server, and receives the planning solution returned by the server. The real-time vehicle information acquisition component communicates with onboard sensors via the AUTOSAR protocol to obtain information such as the vehicle's remaining battery capacity (SOC), location, and energy consumption. The server reporting component processes the data in the ECU into a specified MQTT message format and then sends the processed data via the vehicle's Wi-Fi module. The server downlink data receiving component receives the message via the vehicle's Wi-Fi module and sends it to the ECU via the vehicle's CAN bus for processing into recognizable data.

[0010] The vehicle-mounted ECU module based on the AUTOSAR protocol and the RH850 chip communicates with the information aggregation module through the AUTOSAR RTE standardized communication mechanism and the vehicle-mounted CAN bus to complete data reporting and planning scheme reception, parse the received data, and then feed back the received planning scheme to the user through the vehicle-mounted display human-computer interaction module.

[0011] The human-computer interaction module for proactively pushing charging assistance plans supports destination search. Through subsequent calculations and navigation methods, it provides users with a visual route and a reasonable charging plan. If the user hasn't actively searched for a destination, it can pre-search the destination based on user habits. It also provides regular voice feedback during the interaction process, including voice announcements of key information. It also provides mandatory reminders when the battery level drops below a threshold and supports visual comparison of different planned routes.

[0012] The cloud server supporting information storage and algorithm analysis supports automatic connection to vehicle WIFI, communicates with vehicle data and updates real-time data, collects multi-dimensional data on user charging habits, stores a unique user convenience model for each user, supports the collection of battery-related data during the charging process, and uses an algorithm to establish a battery charging characteristic fitting curve as a characteristic weight for charging time cost calculation.

[0013] The electric vehicle charging path planning method comprises the following steps:

[0014] Step S1: Obtain vehicle battery voltage, current, SOC data and real-time vehicle location, discard outliers, call the multi-dimensional user convenience data and user historical trajectory data stored in the cloud, the multi-dimensional user convenience data includes at least charging time preferences, holiday route habits and charging warning generation habits, integrate and clean the obtained multi-source data to form a structured data set.

[0015] Step S2: Based on the charging characteristic data of the vehicle battery, a constant current-constant voltage segmented charging model is established to fit the battery charging characteristic curve. According to the position of the target power in the fitting curve, the constant current stage charging time cost and the constant voltage stage charging time cost required to reach the target power from the starting power are calculated respectively.

[0016] Step S3: Based on the multidimensional data of user convenience, the hierarchical analysis method in the subjective weighting method is used to determine the preference weights of the user's charging habits and construct a user convenience model.

[0017] Step S4: Generate a reachable charging path set based on the vehicle's remaining mileage and the charging pile location information; combine the user convenience model and the user's historical trajectory data to filter the reachable charging path set, generate a pre-filtered charging path set, and actively push it to the user end.

[0018] In particular, if the user has not set a destination and there is no significant historical pattern (such as random travel on weekends, based on the user's holiday selection habits), the exploration mode is activated and popular areas are recommended based on the remaining mileage. If the battery level falls below the set threshold, the second situation is triggered.

[0019] Step S5: Traverse the pre-screened charging path set and score each path using the TOPSIS method; adjust the score based on real-time traffic information, sort the scores by high and low to generate a scored path set, and feed it back to the human-computer interaction module together with the battery charging cost.

[0020] Step S6: The cloud records the path finally selected by the user, updates the preference weight in the user convenience model based on the user selection, and updates the parameters of the segmented charging model based on the actual charging data to optimize subsequent charging path planning.

[0021] Furthermore, the step S1 includes:

[0022] Step S11: Obtain the vehicle battery voltage, current, SOC data and real-time location of the vehicle through the vehicle information aggregation module.

[0023] Step S12: Using an encrypted communication protocol, the multi-dimensional user convenience data and user historical trajectory data stored in the cloud are retrieved. The multi-source data integration includes collecting user charging time preferences, holiday route habits, charging warning generation habits, and historical trajectory data.

[0024] Step S13: After the structured data set is formed, the desensitized data is encrypted and uploaded to the cloud via the MQTT protocol.

[0025] Furthermore, the step S2 includes:

[0026] Step S21: filtering the charging characteristic data and fitting the segmented charging curves of the constant current stage and the constant voltage stage using the least square method, thereby establishing a constant current-constant voltage segmented charging model based on the charging characteristic data of the vehicle battery.

[0027] Step S22: Calculate the charging time cost of the constant current stage and the constant voltage stage, specifically determine the position of the target power on the fitted segmented charging curve. If it is in the constant current stage, calculate the time from the starting power to the constant current to constant voltage point. If it is in the constant voltage stage, calculate the total time of the constant current stage plus the time from the constant current to constant voltage point to the target power.

[0028] Further, step S3 includes:

[0029] Step S31: constructing a hierarchical structure model based on the multidimensional data of user convenience, and constructing a judgment matrix between elements of each layer.

[0030] Step S32: Perform consistency check on the judgment matrix.

[0031] Step S33: Calculate the weight vector of each layer element, and finally obtain the user charging habit preference weight vector to construct a user convenience model.

[0032] Furthermore, step S4 includes:

[0033] Step S41: The remaining range is determined by collecting the remaining power, combining road condition information and the vehicle's own energy consumption. Based on the remaining range, a reachable charging path set is formed. The reachable charging path set is the set of paths corresponding to all charging pile locations that the vehicle can reach within the remaining range.

[0034] Step S42: A preliminary scoring or screening of accessible paths is performed based on the preference weights in the user convenience model. Further screening is performed based on route preferences or frequently used locations reflected in the user's historical trajectory data to form a pre-screened charging path set. This set is proactively pushed to the user via the human-computer interaction module for selection. If the user selects one of these options, navigation proceeds normally. If not, the process continues to step S5.

[0035] Furthermore, step S5 includes:

[0036] Step S51: Using the determined charging habit preference weights, the TOPSIS method is combined to score and rank different charging paths. The TOPSIS method includes defining multiple evaluation attributes for each path in the pre-screened charging path set, constructing a standardized decision matrix, determining positive and negative ideal solutions, calculating the distance from each path to the positive and negative ideal solutions, and using the normalized scores as the scores.

[0037] Step S52: Obtain a set of scoring paths and feed them back to the human-computer interaction module together with the battery charging cost for the user to make their own selection.

[0038] In order to achieve the above object, the present invention also provides an apparatus for an electric vehicle charging path planning method, comprising:

[0039] A memory for storing a computer program; a processor for implementing a method for planning a charging path for an electric vehicle when executing the computer program;

[0040] An electric vehicle charging route planning system includes a real-time power monitoring module based on a single-chip microcomputer, an information aggregation module connected to all vehicle modules via the CAN bus, an on-board ECU module based on the AUTOSAR protocol and the RH850 chip, a human-computer interaction module that supports the active push of charging assistance plans, and a cloud server that supports information storage and algorithm analysis.

[0041] The beneficial effects of the present invention are:

[0042] The present invention fully considers the role of user charging habits in reasonable charging route planning, explores user charging habits, establishes a user convenience model, and constructs a dynamic time prediction mechanism based on a constant current-constant voltage segmented charging model. By fitting the battery charging curve, the charging time in different power intervals can be accurately calculated. The user's historical trajectory and charging behavior data are integrated, and the hierarchical analysis method is used to establish a dynamic preference weight model to quantify implicit demands such as leisure time utilization and charging pile usage habits. The TOPSIS method is used to combine multiple objective parameters such as real-time road conditions, charging curve delay and user preferences to generate a hierarchical recommendation plan that takes into account both time efficiency and personal convenience. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a flow chart of an electric vehicle charging path planning system provided by the present invention;

[0044] Figure 2 The flow chart for forming and executing the battery charging characteristic fitting curve is presented. DETAILED DESCRIPTION

[0045] The following will be combined with the accompanying drawings to fully and in detail describe the technical solution of the present invention. The specific embodiments are only for a more vivid description of the patent implementation steps and are not all embodiments. Those skilled in the art can obtain other embodiments of the present invention without any creative work, and all of them fall within the scope of protection of the present invention.

[0046] The present invention provides a method for planning a charging path for an electric vehicle, and the specific implementation steps are as follows:

[0047] Step S1: Obtain vehicle battery voltage, current, SOC data and real-time vehicle location, discard outliers, call the multi-dimensional user convenience data and user historical trajectory data stored in the cloud, integrate and clean the obtained multi-source data to form a structured data set.

[0048] The user convenience multi-dimensional data includes at least charging time period preference, holiday route habits and charging warning generation habits. The user convenience multi-dimensional data includes at least charging time period preference, holiday route habits and charging warning generation habits.

[0049] Furthermore, the step S1 includes:

[0050] Step S11: The vehicle's battery voltage, current, SOC data, and real-time vehicle location are acquired through the vehicle information aggregation module. During the hardware data acquisition process, any significant abnormal data is processed. If, for example, the remaining battery SOC is greater than 100%, a data verification mechanism is triggered, the abnormal value is discarded, and a log is recorded.

[0051] Charging characteristic data includes the total battery capacity collected by the real-time power monitoring module , initial remaining power , constant current stage current , constant voltage stage voltage threshold , constant pressure stage time parameters , the transition time from constant current stage to constant voltage stage wait.

[0052] Step S12: Using an encrypted communication protocol, the multi-dimensional user convenience data and user historical trajectory data stored in the cloud are retrieved. The multi-source data integration includes collecting user charging time preferences, holiday route habits, charging warning generation habits, and historical trajectory data.

[0053] User convenience multidimensional data mainly includes holiday route selection habits , charging warning becomes a habit And other multi-dimensional data.

[0054] Furthermore, the step S12 includes:

[0055] Step S121: Basic data of the vehicle user may be collected as charging habit preference parameters for charging habit preference weight evaluation. A preliminary evaluation of the user may be conducted through other forms such as questionnaires.

[0056] In a specific embodiment, the following multi-dimensional data of user convenience are collected:

[0057] User basic information : Including age, gender, occupation, etc., obtained through the form filled out by the user when registering. This information helps to initially understand the user's possible travel and charging patterns. For example, there are differences in working hours and travel patterns among users of different occupations.

[0058] Holiday route selection habits In order to avoid diluting the daily travel route selection habits, holiday route selection should pay attention to identifying whether users have a preference for long-distance travel on holidays. The large flow of people on holidays is often accompanied by staggered charging times, which often has a low overlap with daily travel route selection.

[0059] Charging warning becomes a habit : Extracting features of scenarios where users frequently experience low battery life, combining calendar data to mark workdays, weekends, and holidays, and correlating these with date patterns for low battery events. Specifically, this could include low battery before get off work every Friday, or forgetting to charge after commuting at the end of each month.

[0060] Daily travel route selection habits : Collect users' daily travel time data through the vehicle system or mobile phone application to determine the frequency and pattern of users' travel in different time periods. Use the information aggregation module to record users' frequently traveled routes, departure and destination information, as well as the distribution of charging stations along these routes, to understand users' charging needs when traveling at different distances and upload them to the cloud database. Travel purpose: When users use the vehicle, the application interface prompts users to select the purpose of their trip, such as commuting, shopping, etc. Different travel purposes may lead to different charging behaviors.

[0061] Charging time selection habits : Count the time period of the day when users usually choose to charge, and determine their charging time preferences by analyzing their charging records at different time periods.

[0062] Charging time cost selection habits Time cost is an integral part of charging route planning and is generally divided into driving time, waiting time, and charging time. Here, we emphasize the time cost incurred during the charging process, which is calculated by fitting the battery charging characteristic curve in step S2.

[0063] Step S122: Clean and pre-process the user history trajectory data to form a user history trajectory path set.

[0064] Step S123: Preprocess the user convenience data, focusing on recording the charging time period, recording and analyzing the travel trajectory, and analyzing data such as the environment around the travel destination. This is mainly used to provide feedback on the user's busy and idle time periods in daily life, rationalize user behavior, and record the average time it takes for users to complete their actions. It can also determine whether charging at this time can save overall charging time and other user convenience information.

[0065] Step S13: After forming the structured data set, the desensitized data is encrypted and uploaded to the cloud via the MQTT protocol for use in step S2 to establish the user convenience model.

[0066] Step S2: Based on the charging characteristic data of the vehicle battery, a constant current-constant voltage segmented charging model is established to fit the battery charging characteristic curve. According to the position of the target power in the fitting curve, the constant current stage charging time cost and the constant voltage stage charging time cost required to reach the target power from the starting power are calculated respectively.

[0067] Since battery characteristics such as the battery itself, energy utilization, and battery health will affect the charging process, and the charging characteristics of different batteries will be affected by conditions such as usage time, compared with the traditional method of calculating charging time by power, solving the problem by fitting the battery charging characteristic curve is more accurate.

[0068] Furthermore, the step S2 includes:

[0069] Step S21: filtering the charging characteristic data and fitting the segmented charging curves of the constant current stage and the constant voltage stage using the least square method, thereby establishing a constant current-constant voltage segmented charging model based on the charging characteristic data of the vehicle battery.

[0070] The constant current model :

[0071] .

[0072] in is the function of the remaining power changing with time in the constant current stage, is the initial remaining power, is the constant current, is the total battery capacity, It is the time to switch from constant current stage to constant voltage stage.

[0073] Constant pressure model :

[0074] .

[0075] in is the function of the remaining power changing with time in the constant voltage stage, is the remaining power when switching, is the constant current, is the total battery capacity, The time it takes to switch from the constant current stage to the constant voltage stage, is the time constant of the constant pressure stage.

[0076] make , ensuring the continuity from constant current state to constant voltage state.

[0077] Step S211: sampling the time t and the remaining capacity SOC during the charging process, and filtering the sampled data to remove abnormal values.

[0078] The sampling sequence is , , a sliding average filter is used with a window function length of 5s to eliminate high-frequency noise, and data with adjacent SOC differences exceeding 3% are marked as outliers and removed.

[0079] Step S212: Fit the charging process curve using the least squares method to make the model prediction value and measured values The sum of squared distances is minimized.

[0080] .

[0081] The constraint parameters are .

[0082] Step S22: Calculate the charging time cost of the constant current stage and the constant voltage stage, specifically determine the position of the target power on the fitted segmented charging curve. If it is in the constant current stage, calculate the time from the starting power to the target power. If it is in the constant voltage stage, calculate the total time of the constant current stage plus the time from the constant current to constant voltage point to the target power.

[0083] Furthermore, the step S22 includes:

[0084] Step S221: Using the single chip microcomputer-based power real-time monitoring module to monitor the remaining power at the start of charging , end charging If the user does not enter any information, the default value is to fill the space directly. Set to 100%. In actual application scenarios, users can manually enter the desired charging level.

[0085] Step S222: Determine the position of the fitting curve where the charging power ends. Specifically, there are three situations:

[0086] In the first case, if , the whole process is in constant current state;

[0087] In the second case, if , then the charging process includes the conversion process from constant current to constant voltage;

[0088] In the third case, if , the whole process is in a constant pressure state.

[0089] Step S223: Bring in the constant current and constant voltage model of charging to obtain the time cost of the constant current process , time cost of constant pressure process .

[0090] in ,

[0091] .

[0092] Step S224: Obtain the corresponding charging time cost based on the judgment condition in step S42.

[0093] .

[0094] in The cost is the charging time.

[0095] Step S3: Based on the multidimensional data of user convenience, the hierarchical analysis method in the subjective weighting method is used to determine the preference weights of the user's charging habits and construct a user convenience model.

[0096] Furthermore, step S3 includes:

[0097] Step S31: constructing a hierarchical structure model based on the multidimensional data of user convenience, and constructing a judgment matrix between elements of each layer.

[0098] Specifically, according to six different charging habit preference parameter sets: , the following judgment matrix can be established:

[0099] .

[0100] in Representation and charging habit preference parameters compared to, Therefore, and .when When the two indicators are the same, they must be equally important, so the main diagonal elements are all 1. After investigation and data integration, it is found that compared with the basic information of users In terms of the user's charging habit profile, other influencing factors have a greater impact on the user's charging habit profile. Therefore, when actually testing the authenticity and reliability of the matrix, the first column elements can be used as one of the bases for preliminary judgment, that is, To meet the actual situation.

[0101] Step S32: Perform consistency check on the judgment matrix:

[0102] Calculating consistency index ,in is the number of charging habit preference parameters;

[0103] Find the corresponding average random consistency index :

[0104]

[0105] Calculate the consistency ratio .

[0106] if , then the consistency of the judgment matrix can be considered acceptable, otherwise the judgment matrix needs to be revised and analyzed.

[0107] Step S33: Then use the first column of the judgment matrix to use The weight vector of each layer element can be calculated, and finally the user charging habit preference weight vector is obtained to build a user convenience model.

[0108] The calculation of charging habit preference weights here is primarily based on the habits of different users, so the subjective hierarchical analysis method (AHP) has better applicability and accuracy. Alternatively, the entropy weight method can be used to calculate the objective charging habit preference weights for each indicator. The charging habit preference weights are determined based on the degree of dispersion of each indicator's data. The greater the degree of dispersion, the greater the indicator's ability to distinguish between options, and the corresponding charging habit preference weight increases.

[0109] Step S4: Generate a reachable charging path set based on the vehicle's remaining mileage and the charging pile location information; combine the user convenience model and the user's historical trajectory data to filter the reachable charging path set, generate a pre-filtered charging path set, and actively push it to the user end.

[0110] Step S41: The remaining range is determined by collecting the remaining power, combining road condition information and the vehicle's own energy consumption. Based on the remaining range, a reachable charging path set is formed. The reachable charging path set is the set of paths corresponding to all charging pile locations that the vehicle can reach within the remaining range.

[0111] Furthermore, step S41 includes:

[0112] In a specific embodiment, the recommendation scheme based on the remaining mileage is mainly divided into the following situations.

[0113] In the first case, the user actively enters the destination. Compare the remaining mileage with the required mileage. If the remaining mileage is greater than the mileage required to reach the destination, When Indicates the distance required to reach the destination. Indicates the mileage required to reach the destination, k indicates the safety factor, which is 1.2 by default (reserving 20% ​​redundant power). The owner can preset the route planning based on the actual situation, and no charging prompt is given. If the remaining mileage is less than the required mileage, that is When it is displayed, it means that there may be risks in arriving at the destination, and a multi-stop route planning is performed.

[0114] The multi-stop route planning is based on As a principle, the remaining mileage is less than the required mileage for charging. N charging stations are inserted in the total route to ensure that the battery can be replenished on the way to reach the final destination. At the last charging station before the foreseeable destination, the Principles, among which The mileage corresponds to the power threshold, and the vehicle must reach its destination without exceeding the mileage corresponding to the power threshold.

[0115] In the second case, if the remaining mileage is less than the set threshold, that is, When a charging decision is triggered, the system will prioritize finding the nearest charging station, regardless of whether the destination is known, and re-plan the route to ensure the vehicle does not stop midway due to battery depletion. In this case, even if the user has a clear destination, the battery issue should be resolved first.

[0116] In the third case, the user has not set a destination, but the system detects that the current time / location is highly consistent with the user's historical trajectory path set. If the remaining power can cover the predicted path and there are available charging facilities near the area, the system will make charging suggestions based on intelligent judgment and recommend charging in advance.

[0117] In the fourth case, if the user has not set a destination and there is no significant historical pattern (such as random travel on weekends, based on the user's holiday travel habits), the exploration mode is activated and popular areas are recommended based on the remaining mileage. The second case is triggered if the battery level falls below the set threshold.

[0118] Step S42: A preliminary scoring or screening of accessible paths is performed based on the preference weights in the user convenience model. Further screening is performed based on route preferences or frequently used locations reflected in the user's historical trajectory data to form a pre-screened charging path set. This set is proactively pushed to the user via the human-computer interaction module for selection. If the user selects one of these options, navigation proceeds normally. If not, the process continues to step S5.

[0119] Step S5: Traverse the pre-screened charging path set and score each path using the TOPSIS method; adjust the score based on real-time traffic information, sort the scores by high and low to generate a scored path set, and feed it back to the human-computer interaction module together with the battery charging cost.

[0120] Furthermore, step S5 includes:

[0121] Step S51: Using the determined charging habit preference weights and the TOPSIS method, different charging paths are ranked and scored. Based on each path's performance on various metrics, its distance from the ideal and negative ideal charging paths is calculated, resulting in a relative proximity between each path. This serves as a basis for evaluating the path's quality and providing the user with the optimal charging path recommendation.

[0122] In a specific embodiment, the scores are normalized to obtain .

[0123] Among them The unnormalized score of the route is .

[0124] Among them The distance between the route and the negative ideal solution is .

[0125] No. The distance between the route and the ideal solution is .in is the proportion of each charging habit preference parameter in step S3.

[0126] The negative ideal solution is .

[0127] The positive ideal solution is .

[0128] The set of all routes that meet the user's travel purpose is ,Establish The forward matrix consisting of all possible path solutions and 6 influencing factors:

[0129] .

[0130] Then the standardized decision matrix is ​​recorded as , Each element in is .

[0131] Step S52: Evaluate all path plans under the various factors that may affect the charging path score coefficients that the user may encounter during this trip, and derive path optimization plans that meet different user habits in the form of substantive scores from large to small. Obtain a scored path set and feed it back to the human-computer interaction module together with the battery charging cost for the user to make their own selection.

[0132] Step S6: The cloud records the path finally selected by the user, updates the preference weight in the user convenience model based on the user selection, and updates the parameters of the segmented charging model based on the actual charging data to optimize subsequent charging path planning.

[0133] To achieve the above objectives, the present invention also provides an apparatus for an electric vehicle charging route planning method, comprising: a memory for storing a computer program; a processor for implementing an electric vehicle charging route planning method when executing the computer program. A system for an electric vehicle charging route planning method includes a real-time power monitoring module based on a single-chip microcomputer, an information aggregation module connected to all vehicle modules via a CAN bus, an on-board ECU module based on the AUTOSAR protocol and the RH850 chip, a human-computer interaction module that supports the active push of charging assistance plans, and a cloud server that supports information storage and algorithm analysis.

[0134] The specific embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should fall within the scope of protection determined by the claims of the present invention.

Claims

1. A method for planning a charging path for an electric vehicle, characterized in that: The following steps are involved: S1. Obtain vehicle battery voltage, current, SOC data, and real-time vehicle location, discard outliers, and access multi-dimensional user convenience data and historical user trajectory data stored in the cloud. The multi-dimensional user convenience data includes at least charging time preferences, holiday route habits, and charging warning generation habits. The obtained multi-source data is integrated and cleaned to form a structured dataset. S2. Based on the charging characteristic data of the vehicle battery, establish a constant current-constant voltage segmented charging model to fit the battery charging characteristic curve. Based on the position of the target power in the fitting curve, calculate the constant current stage charging time cost and the constant voltage stage charging time cost required to reach the target power from the starting power; S3. Based on the multidimensional data of user convenience, the user charging habit preference weights are determined using the analytic hierarchy process in the subjective weighting method to construct a user convenience model; S4. Generate a set of accessible charging paths based on the vehicle's remaining mileage and charging station location information; Combining the user convenience model and the user's historical trajectory data, the reachable charging path set is screened, a pre-screened charging path set is generated, and the pre-screened charging path set is actively pushed to the user end; S5. Traverse the pre-screened charging path set and score each path using the TOPSIS method; adjust the score based on real-time traffic information, generate a score path set by ranking, and feed it back to the human-computer interaction module; S6. The cloud records the path finally selected by the user, updates the preference weight in the user convenience model based on the user selection, and updates the parameters of the segmented charging model based on the actual charging data to optimize subsequent charging path planning.

2. The electric vehicle charging path planning method according to claim 1, characterized in that: The step S1: The vehicle battery voltage, current, SOC data and real-time location of the vehicle are obtained through the on-board information aggregation module; the encrypted communication protocol is used to call the multi-dimensional user convenience data and user historical trajectory data stored in the cloud; multi-source data integration includes collecting user charging time preferences, holiday route habits, charging warning generation habits and historical trajectory data; after forming a structured data set, the desensitized data is encrypted and uploaded to the cloud through the MQTT protocol.

3. The electric vehicle charging path planning method according to claim 1, characterized in that: Step S2: The charging characteristic data is filtered and fitted using the least squares method to obtain the segmented charging curves for the constant current stage and the constant voltage stage, forming a constant current-constant voltage segmented charging model based on the charging characteristic data of the vehicle battery; the charging time cost for the constant current stage and the constant voltage stage is calculated, specifically by determining the position of the target power on the fitted segmented charging curve. If it is in the constant current stage, the time from the starting power to the target power is calculated; if it is in the constant voltage stage, the total time of the constant current stage plus the time from the constant current to constant voltage point to the target power is calculated.

4. The electric vehicle charging path planning method according to claim 1, characterized in that: The method of using the hierarchical analysis method in the subjective weighting method to determine the user's charging habit preference weight includes: A hierarchical model is constructed based on the multidimensional data of user convenience, a judgment matrix between elements of each layer is constructed, and a consistency test is performed on the judgment matrix. After passing the consistency test, the weight vector of the elements of each layer is calculated, and finally the user charging habit preference weight vector is obtained to construct a user convenience model.

5. The electric vehicle charging path planning method according to claim 1, characterized in that: In the step S4: The reachability charging path set refers to the set of paths corresponding to all charging pile locations that the vehicle can reach within the remaining mileage; Screening based on the user convenience model and user historical trajectory data means preliminarily scoring or screening the accessible paths based on the preference weights in the user convenience model, and further screening based on the route preferences or frequently used locations reflected in the user historical trajectory data to form a pre-screened charging path set.

6. The electric vehicle charging path planning method according to claim 1, characterized in that: The TOPSIS method is used to score each path, including: A plurality of evaluation attributes are defined for each path in the pre-screened charging path set, a standardized decision matrix is ​​constructed, a positive ideal solution and a negative ideal solution are determined, a distance from each path to the positive ideal solution and the negative ideal solution is calculated, and the normalized score is used as a score.

7. A device for implementing the electric vehicle charging path planning method according to any one of claims 1 to 6, characterized in that: The device comprises: The real-time power monitoring module based on the single-chip microcomputer is designed and integrated into the vehicle battery management system (BMS) to collect battery voltage, current and SOC data in real time; An information aggregation module that connects all vehicle modules based on the CAN bus, implemented based on the CAN bus and AUTOSAR protocol, connects the vehicle's electronic control units ECU, and is used to aggregate the vehicle's real-time location information and the data from the power real-time monitoring module, and upload it after preliminary pre-processing; A human-computer interaction module for actively pushing the charging assistance plan is used to actively push the pre-screened charging path set and the scored path set to the user, and receive the user's path selection result and real-time road condition feedback information input by the user; A cloud server supporting information storage and algorithm analysis is used to store preprocessed vehicle data, user convenience multidimensional data, user historical trajectory data, segmented charging model parameters and user convenience model; performs fitting of the constant current-constant voltage segmented charging model and charging time cost calculation; performs the hierarchical analysis method to analyze the user convenience multidimensional data to construct and update the user convenience model, the analysis paying special attention to holiday route habits and charging warning generation habit data; performs the generation algorithm of the reachability charging path set and the pre-screened charging path set; performs the TOPSIS method to score and rank the paths; and performs the update optimization algorithm of the user convenience model and segmented charging model parameters.

Images