Intelligent reservation charging method and device for extended-range plug-in hybrid electric vehicle, and computer system

Through the multi-wheel recommendation method, combining the historical data, map information and real-time road conditions of the vehicle and charging station, the reservation charging of extended-range plug-in hybrid vehicles is optimized, solving the problems of waste of resources and poor user experience in the existing technology. A more economical and reliable charging solution is achieved.

WO2024239494A1PCT designated stage expired Publication Date: 2024-11-28DONGFENG MOTOR GRP

Patent Information

Application Number
PCT/CN2023/118295
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-25
Filing Date
2023-09-12
Publication Date
2024-11-28

Smart Images

  • Figure CN2023118295_28112024_PF_FP_ABST
    Figure CN2023118295_28112024_PF_FP_ABST
Patent Text Reader

Abstract

An intelligent reservation charging method for an extended-range plug-in hybrid electric vehicle, comprising the following steps: recommendation for the first round: selecting a specific method to choose a first batch of recommended charging stations, predicting the number of idle fast and slow charging stations and the total number of idle charging stations at a reservation charging time point, and choosing a finally recommended reservation charging station for the first round; recommendation for the second round: by combining map information and power consumption calculation logic in a pure electric mode, calculating the remaining SOC allowing for arriving at the finally recommended reservation charging station for the first round, and choosing a recommended reservation charging station for the second round; and final recommendation: when the reservation time point is reached, recommending a final charging station according to real-time traffic information on a map and the power consumption calculation logic in the pure electric mode and according to the SOC and real-time information of the idle fast and slow charging stations. The present invention also relates to an intelligent reservation charging device for an extended-range plug-in hybrid electric vehicle, and a computer system. The present invention has the function of arranging an optimal reservation charging station and an optimal route according to user's reservation conditions, and can be widely applied to extended-range plug-in hybrid electric vehicles.
Need to check novelty before this filing date? Find Prior Art

Description

An intelligent method, device and computer system for pre-charging extended-range plug-in hybrid electric vehicle Technical Field

[0001] The present invention relates to an extended-range plug-in hybrid electric vehicle, and in particular to an intelligent method, device and computer system for pre-charging an extended-range plug-in hybrid electric vehicle. Background Art

[0002] Extended-range plug-in hybrid electric vehicles (PHEVs) can effectively address national energy conservation and emission reduction requirements and alleviate customers' range anxiety. Combining a traditional internal combustion engine with a pure electric vehicle motor, they can reduce dependence on and demand for petroleum fuels, improving the vehicle's fuel economy and emissions performance. PHEVs typically feature the following driving modes: 1. Pure electric mode, in which the vehicle relies on a high-voltage battery to power the drive motor, which then drives the vehicle. 2. Series mode, in which the engine runs, driving a generator to charge the high-voltage battery and then to power the drive motor. Therefore, when the battery charge is low, the series mode can be used to drive the vehicle, eliminating customers' range anxiety.

[0003] With regard to new energy vehicles, the industry has noticed their charging needs and has tried to propose relevant solutions. For example, the State Grid Corporation of China has an invention application with the publication number CN104779680A and the invention name "Electric Vehicle Reservation Charging Control Method and Device", which discloses an electric vehicle reservation charging control method and device. Among them, the method includes: obtaining reservation charging information and pre-set optimization control parameters; obtaining vehicle status information of the vehicle pre-bound to the reservation charging information; determining a set of reservation charging plans for the vehicle based on the reservation charging information and the vehicle status information; and using the optimization control parameters to filter out a recommended charging plan from the reservation charging plan set. The present invention solves technical problems such as the waste of charging equipment resources and poor user charging experience caused by the inability to optimize the reservation charging of electric vehicles.

[0004] However, the invention also has the following shortcomings: 1. The invention patent application is for pure electric vehicles, while the present invention is for extended-range plug-in hybrid vehicles, which have significant differences in the overall vehicle mode; 2. The driver's specific driving habits are not taken into account, and the driver's actual driving mode and habitual user behavior analysis is not conducted; 3. It is mainly aimed at the grid side or the charging pile side, and no specific description is given for the specific vehicle control parameters; 4. The impact of actual road conditions on scheduled charging is not considered.

[0005] Another utility model, filed by Jiangsu Jintan Damai Automotive Engineering Research Institute Co., Ltd. and Jiangsu Jintan Automobile Industry Co., Ltd., with publication number CN207683369U and the name of the utility model, entitled "A Vehicle Pre-booked Charging System," discloses a vehicle pre-booked charging system comprising a remote control terminal, a TBOX, an onboard charger (OBC), a battery management system (BMS), and a vehicle controller (VCU). The remote control terminal is connected to the TBOX, which is connected to the onboard charger (OBC) and the battery management system (BMS), respectively. The battery management system (BMS) is connected to the vehicle controller (VCU). The advantages of this utility model are: direct network management controls the shutdown and start / stop of the BMS and TBOX, thereby simplifying the implementation of the pre-booked charging function; remotely setting the scheduled charging time via a mobile phone facilitates and simplifies the charging process. Similar to the previous invention, its shortcomings are: 1. It does not distinguish between specific new energy vehicle models and does not provide a detailed description of the vehicle mode and control; 2. It does not take into account the user's previous charging habits, the conditions of the charging station itself, and actual road conditions.

[0006] Summary of the Invention

[0007] The purpose of the present invention is to overcome the shortcomings of the above-mentioned background technology and provide an intelligent method, device and computer system for pre-booking charging of extended-range plug-in hybrid electric vehicles, so that it has the function of reasonably arranging the optimal reserved charging station and route information according to the user's reservation conditions.

[0008] The present invention provides an intelligent method for charging reservations for extended-range plug-in hybrid electric vehicles, which performs multiple rounds of recommendations for reserved charging stations, including the following steps: first-round recommendation: selecting a first batch of recommended charging stations based on the number of reserved charging records, predicting the number of idle fast and slow charging stations and the total number of idle charging stations at the reserved charging time based on the historical idle data of the charging stations, and selecting the final recommended reserved charging stations in the first round based on the idle number threshold; second-round recommendation: calculating the remaining SOC when reaching the final recommended reserved charging station in the first round based on map information and the power consumption calculation logic in pure electric mode, and recommending the second round of recommended reserved charging stations based on the remaining SOC; final recommendation: when arriving at the reserved time point, calculating the SOC when reaching the second round of recommended reserved charging stations based on the real-time traffic information on the map and the power consumption calculation logic in pure electric mode, and making the final charging station recommendation based on the SOC and the real-time fast and slow charging idle information of the charging stations.

[0009] In the above technical solution, the specific process of the first round of recommendation steps is as follows: departure information reservation: input the departure location and departure time of the scheduled charging through the vehicle computer or mobile phone; first reservation record acquisition: if the system has less than n user reservation charging records, the system calculates the y nearest charging stations within x km near the starting point as recommended charging stations, and obtains the charging station information; second reservation record acquisition: if the system has more than n user reservation charging records, first sort the charging stations actually used by the user by the number of times used, select the z charging stations with the most uses, calculate the distance between the departure point and the selected charging station, and if it is less than or equal to the distance threshold, retain it as the recommended charging station, and obtain the charging station information; if it is greater than the distance threshold, remove it; information extraction: extract the information on the number of idle charging guns of the recommended charging station in the past m days from the charging station history data system.

[0010] In the above technical solution, the information extraction step includes the following processes: data set: the data on the number of idle fast and slow charging guns at the recommended charging station for m days are divided into an input set and a target set, the data of the first m1 days are used as the input data set, and the data of the last m2 days are used as the target data set, wherein m=m1+m2, and m1>m2; fast and slow charging empty value filling: if there is an empty value in the fast and slow charging data of a station for m1 days, the fast charging empty value is filled with the average value μ1 of the idle number of fast charging in m1 days, and the slow charging empty value is filled with the average value μ2 of the idle number of slow charging; normalization: the fast and slow charging data of m1 and m2 days are normalized respectively; time series analysis: the time series machine learning model is used for time series analysis, and the root mean square error R is used. 2 To evaluate time series models, choose R 2 The model with the smallest value predicts the number of fast and slow charging stations available at a certain charging station at the driver's reservation time and recommends it as the final charging station in the first round.

[0011] In the above technical solution, the normalization step is the min-max method, and the normalized fast charging idle number is: (fast – fast min ) / (fast max –fast min ), Similarly, the normalized slow charging idle number is: (slow–slow fast ) / (slow max –slow min ), similarly, use the average value of the fast charge number μ3 and the average value of the slow charge number μ4 of the target set m2 to fill the empty values ​​of the target data set, and use the maximum value of the target set fastTarget max and fastTarget min Denormalize the target set.

[0012] In the above technical solution, in the time series analysis step, the process of predicting the number of fast and slow charging idle stations at a charging station at the driver's reservation time is: if a station finally predicts that the number of fast charging idle stations is greater than fast0 and the number of slow charging idle stations is greater than slow0, and the total number of idle stations that is the sum of the fast charging idle number and the slow charging idle number is greater than total0, then it is recommended as the final charging station in the first round.

[0013] In the above technical solution, in the time series analysis step, the time series machine learning model includes but is not limited to the LSTM\RNN model.

[0014] In the above technical solution, the specific process of the second round of recommendation steps is as follows: map information acquisition: according to the starting point input by the user and the charging stations recommended in the first round, the map information from the starting point to each charging station is obtained; speed segment map conversion: according to the actual map road condition information, the road condition information is converted into a speed segment map; remaining SOC calculation: according to the actual SOC of the vehicle, combined with the pure electric mode cruising range calibration table, the actual power consumption is obtained, and the remaining SOC1 after the vehicle completely passes through the first section of the road can be calculated; the remaining SOCn after passing through the second section of the road, the third section of the road... and even the nth end can be calculated in sequence; pure electric mode selection: if the SOC is less than 0 after a certain section of the road, it means that the pure electric mode cannot reach the recommended charging station, and finally the charging station with a remaining SOCn>0, that is, the charging station that can be reached by the pure electric mode, is selected as the second round of recommended charging stations, and the accessible route information is provided; series mode selection: if the final SOCn<=0, the user is prompted that the series mode is required to reach some of the charging stations in the first round of recommended charging stations used for the second round of recommendations.

[0015] In the above technical solution, the specific process of the final recommendation step is as follows: Pure electric mode SOC calculation: When the user's scheduled time point is reached, the actual map information of the route to the second round of recommended charging stations at that time point is obtained from the map, and the remaining SOC when the pure electric mode reaches the second round of recommended charging stations is recalculated according to the two-round recommendation steps. If the calculated remaining SOC is <0, the user is prompted to use the series mode to reach the station; Comparison and sorting recommendations: If the remaining SOC>0, it is used as a recommended charging station, and the number of fast charging idle stations, slow charging idle stations and total idle stations of the recommended charging station are queried in the charging station information system. If the fast charging idle station is greater than fastLast, the slow charging idle station is greater than slowLast and the total idle station is greater than totalLast, it can be used as the final recommended charging station. The final recommended charging stations are sorted from large to small according to the remaining SOC as the final recommended charging station list, and the corresponding driving routes are recommended.

[0016] The present invention further provides an intelligent device for pre-charging an extended-range plug-in hybrid electric vehicle, which stores a computer program capable of executing an intelligent method for pre-charging an extended-range plug-in hybrid electric vehicle.

[0017] The present invention also provides a computer system comprising an intelligent device for pre-charging an extended-range plug-in hybrid electric vehicle.

[0018] The present invention provides an intelligent method, device, and computer system for scheduled charging of an extended-range plug-in hybrid electric vehicle, which has the following beneficial effects:

[0019] The present invention is mainly aimed at extended-range plug-in hybrid electric vehicles. Taking into account actual road conditions, user history, the actual conditions of the vehicle itself and charging stations, and based on the advantages of cloud platforms such as low cost, reliability and good scalability, an intelligent system and method for extended-range plug-in hybrid electric vehicle charging reservation is established on the cloud platform. The system considers vehicle characteristics, user history charging behavior, road conditions for the vehicle to the charging station, and charging station information, etc., and can provide extended-range plug-in hybrid electric vehicle users with the most economical and available reserved charging stations and driving routes. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] FIG1 is a schematic diagram of the power structure of a range-extended plug-in hybrid electric vehicle according to the present invention;

[0021] FIG2 is a schematic diagram of the overall flow of the intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to the present invention;

[0022] FIG3 is a schematic diagram of a specific flow chart of the first round of recommendation steps in the intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to the present invention;

[0023] FIG4 is a schematic diagram of a specific flow chart of the second round of recommendation steps in the intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to the present invention;

[0024] FIG5 is a diagram of actual road speed segments in sub-step 2.2 of the second-round recommendation step in the intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to the present invention;

[0025] FIG6 is a schematic diagram of a specific flow chart of the final recommendation step in the intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to the present invention;

[0026] FIG7 is a schematic structural diagram of an intelligent device for scheduled charging of a range-extended plug-in hybrid electric vehicle according to the present invention;

[0027] FIG8 is a schematic diagram of the structure of the computer system of the present invention. DETAILED DESCRIPTION

[0028] The present invention will be further described in detail below with reference to the accompanying drawings and examples, but the examples should not be construed as limiting the present invention.

[0029] The power structure of the extended-range plug-in hybrid electric vehicle (PHEV) described herein is shown in Figure 1 , where the engine is a gasoline engine. Solid lines in the figure represent mechanical connections, while dashed lines represent electrical connections. There are two drive modes: electric mode and series mode. In electric mode, the engine is inoperative, and the high-voltage battery provides power to the drive motor, which in turn drives the vehicle. In series mode, the engine operates, driving the generator to charge the high-voltage battery and the generator to provide power to the drive motor, which in turn drives the vehicle.

[0030] Referring to Figure 2 , the present invention's intelligent method for pre-booking charging for extended-range plug-in hybrid electric vehicles is implemented within a server, vehicle computer, or controller, and is connected to a client within the vehicle computer or mobile terminal. The server, vehicle computer, or controller utilizes the Internet of Things to obtain charging station information and map information, and also obtains information such as the vehicle's remaining SOC from the vehicle. The client within the vehicle computer or mobile terminal allows the user to enter the departure location and time for a scheduled charging appointment and displays a map and charging station location information. In this embodiment, the server is network-connected to a charging station information system and retrieves various charging station parameter information from the charging station information system. The charging station information system can be an existing external system or an integral part of the present invention.

[0031] The overall idea of ​​the present invention is as follows:

[0032] First-round recommendations: Based on the number of scheduled charging records, a specific method is used to select the first batch of recommended charging stations. Based on the historical idle data of charging stations, the number of idle fast and slow charging stations and the total idle number of charging stations at the scheduled charging time are predicted. Based on the idle number threshold, the final recommended charging stations for the first round are selected.

[0033] Referring to Figure 3, the logic diagram of the recommended charging stations for the first round of scheduled charging is as follows:

[0034] 1.1. The driver enters the departure location and time for the scheduled charging appointment via the vehicle computer or mobile phone;

[0035] If the number of user charging reservations in the system is less than n (e.g., 500, configurable), the system calculates the y nearest charging stations (configurable, e.g., 40) within x km (configurable, e.g., 20 km) of the starting point as recommended charging stations and obtains the charging station information.

[0036] 1.3. If the system has more than n (e.g., 500, configurable) user charging reservation records, the system first sorts the user's actual charging stations by their frequency of use. The z (e.g., 50, configurable) most frequently used charging stations are selected. The distance between the departure point and the selected charging station is calculated. If the distance is less than or equal to a configurable distance threshold (e.g., 20km), the station is retained as a recommended charging station and the charging station information is obtained. If the distance is greater than the threshold, the station is removed.

[0037] 1.4. Extract the number of idle charging guns at the recommended charging station over the past 30 days from the charging station historical data system;

[0038] 1.4.1. The sampling frequency of charging station idle information in the charging station information system is t hours (configurable, default 1 hour). The 30-day data on the number of idle fast and slow charging guns at recommended charging stations is processed as follows: the 30-day data on the number of idle fast and slow charging guns at recommended charging stations is divided into an input set and a target set. The data for the first 27 days is used as the input data set, and the data for the last 3 days is used as the target data set.

[0039] 1.4.2. If there are empty values ​​in the 27-day fast and slow charging data of a station, the 27-day average values ​​of the idle fast and slow charging data μ1 (corresponding to the fast charging average) and μ2 (corresponding to the slow charging average) are used to fill the empty fast and slow charging values ​​respectively;

[0040] 1.4.3. Normalize the 27-day fast and slow charging data separately (min-max method or z-score method):

[0041] 1.4.3.1. Taking the min-max method as an example, the normalized fast-charging idle count is: (fast – fastmin) / (fastmax – fastmin). Similarly, the normalized slow-charging idle count is: (slow – slowfast) / (slowmax – slowmin). Similarly, the average values ​​of the fast and slow charging counts μ3 and μ4 of the target set (3 days of data) are used to fill the empty values ​​in the target data set (corresponding to the last 3 days of data). The maximum values ​​of the target set, fastTargetmax and fastTargetmin, are used to normalize the target set.

[0042] 1.4.4. Use time series machine learning models (not limited to LSTM\RNN models) for time series analysis, using the root mean square error R 2 To evaluate time series models, choose R 2 The model with the smallest value predicts the number of fast and slow charging idle cells at a charging station at the driver's reservation time (accurate to h).

[0043] 1.4.4.1. If a station's final predicted number of fast-charging idle stations is greater than fast0 (configurable, for example, 2), its number of slow-charging idle stations is greater than slow0 (configurable, for example, 2), and its total idle station number (the sum of the fast-charging and slow-charging idle stations) is greater than total0 (configurable, for example, 2), it will be recommended as the final charging station in the first round.

[0044] Second round of recommendations: Based on map information and the power consumption calculation logic in pure electric mode, the remaining SOC at the final recommended charging station in the first round is calculated, and the second round of recommended charging stations is recommended based on the remaining SOC.

[0045] Referring to Figure 4, the following method is used to recommend charging stations for the second round of reservations:

[0046] First is the technical principle analysis of calculating the remaining SOC of the car: in pure electric mode, the calculation logic of the actual pure electric remaining SOC is disclosed in detail in Figure 4. Table 1 is the basic calibration table of battery consumption in pure electric mode, where the horizontal axis is the remaining SOC of the high-voltage battery, and the vertical axis is the uniform vehicle speed. The data in the table represents the percentage of power consumption per kilometer when the horizontal axis SOC reaches 0% when the vehicle speed is equal to the vertical axis speed; Table 2 represents the ambient temperature correction coefficient, which corrects the power consumption in Table 1; Table 3 represents the power correction coefficient, which corrects Table 1 according to the user's past historical behavior. The specific historical behavior analysis is as follows: 1) When the vehicle's driving distance is less than or equal to 1000km, the default power correction table is used. is a power correction table, as shown in the values ​​in Table 3; 2) When the vehicle travel distance is greater than 1000km, based on the user's historical behavior, the power at a certain speed (10km / h, 20km / h, 30km / h, 40km / h, 50km / h, 60km / h, 70km / h, 80km / h, 100km / h, 120km / h, 140km / h, 160km / h, etc.) in the user's history is counted, and the average value is calculated as the driving demand power at that speed, and divided by the demand power under uniform speed conditions to obtain the power correction coefficient at that speed. The demand power under uniform speed can be obtained through actual road tests. If this speed is not available in the historical behavior data, the default value in Table 3 is used.

[0047] Table 1 Basic calibration table of pure electric mode driving range

[0048] Table 2 Ambient temperature correction table

[0049] Table 3 Power correction table

[0050] 2.1. Based on the user's input starting point and the first round of recommended charging stations, obtain map information from the starting point to each charging station (multiple driving routes are possible);

[0051] 2.2. Taking a recommended charging station as an example, based on actual map traffic information (such as vehicle speed), the traffic information is converted into a speed segment diagram, as shown in Figure 5. In this diagram, the maximum speed of the first road segment is v1. If the vehicle's mileage is less than 50 km, the vehicle's speed for this segment is set to v1. If the vehicle's mileage is ≥ 50 km, the average speed of the customer's historical travel through this speed segment is used as the vehicle's speed for this segment.

[0052] 2.3. Based on the vehicle's actual SOC and the pure electric mode range calibration table, the actual power consumption can be calculated. The remaining SOC1 after the vehicle has completely completed the first section of the road can be calculated. The remaining SOCn after the vehicle has completed the second section, the third section, and so on can be calculated.

[0053] 2.4. If the SOC is less than 0 after a certain distance m, it indicates that the recommended charging station cannot be reached in pure electric mode. The charging station with a remaining SOCn>0 (i.e., reachable in pure electric mode) is selected as the second round of recommended charging stations, and the reachable route information is provided;

[0054] 2.5. If the final SOCn is less than or equal to 0, the user is prompted to use the series mode (with the engine running) to reach the portion of the charging stations recommended in the first round that will be used for the second round of recommendations.

[0055] 3. Final recommendation: When the scheduled time arrives, the SOC of the charging station recommended for the second round of reservations is calculated based on the real-time traffic information on the map and the power consumption calculation logic of the pure electric mode. The final charging station recommendation is made based on the SOC and the real-time fast and slow charging idle information of the charging station.

[0056] See Figure 6, 3.1. When the user's scheduled time arrives, the actual map information of the route to the second round of recommended charging stations at that time is obtained from the map. The remaining SOC when reaching the second round of recommended charging stations in pure electric mode is recalculated according to step 4. If the calculated remaining SOC is less than 0, the user is prompted to use series mode to reach the station;

[0057] 3.2. If the remaining SOC is greater than 0, it is recommended as a charging station. In the charging station information system, query the number of fast charging idle stations, slow charging idle stations, and total idle stations for the recommended charging station. If the fast charging idle station number is greater than fastLast (which can be set, for example, 2), the slow charging idle station number is greater than slowLast (which can be set, for example, 2), and the total idle station number is greater than totalLast (which can be set, for example, 2), it is selected as the final recommended charging station. The final recommended charging stations are sorted from highest to lowest remaining SOC, forming the final recommended charging station list and the corresponding driving route.

[0058] 7 , the intelligent device for pre-booking charging of a range-extended plug-in hybrid electric vehicle of the present invention performs multiple rounds of recommendations for pre-booked charging stations, including the following contents:

[0059] First-round recommendation module: Based on the number of scheduled charging records, a specific method is selected to select the first batch of recommended charging stations. Based on the historical idle data of charging stations, the idle number of fast and slow charging stations and the total idle number of charging stations at the scheduled charging time are predicted. Based on the idle number threshold, the final recommended charging stations for the first round are selected;

[0060] Second-round recommendation module: Based on map information and the power consumption calculation logic in pure electric mode, it calculates the remaining SOC when reaching the final recommended charging station in the first round, and recommends the second-round recommended charging station based on the remaining SOC;

[0061] Final recommendation module: When the appointment time is reached, the SOC of the charging station recommended for the second round of appointments is calculated based on the real-time traffic information on the map and the power consumption calculation logic of the pure electric mode. The final charging station recommendation is made based on the SOC and the real-time fast and slow charging idle information of the charging station.

[0062] 8 , the computer system of the present invention includes an intelligent device for pre-charging an extended-range plug-in hybrid electric vehicle.

[0063] The technical principles and key points of the present invention are:

[0064] An intelligent method and device for charging reservations for extended-range plug-in hybrid electric vehicles. The method makes multiple rounds of recommendations for reserved charging stations. Based on the number of reserved charging records, a specific method is selected to select the first batch of recommended charging stations. Based on the historical idle data of the charging stations, the idle fast and slow charging idle numbers and the total idle number of charging stations at the reserved charging time are predicted, and based on the idle number threshold, the final recommended reserved charging stations in the first round are selected. Based on the map information and the power consumption calculation logic in pure electric mode, the remaining SOC when reaching the final recommended reserved charging station in the first round is calculated, and the second round of recommended reserved charging stations is recommended based on the remaining SOC. When the reserved time point is reached, the SOC when reaching the second round of recommended reserved charging stations is calculated based on the real-time traffic information on the map and the power consumption calculation logic in pure electric mode, and the final charging station recommendation is made based on the SOC and the real-time fast and slow charging idle information of the charging stations. Among them,

[0065] 1. As described above, the specific method for recommending the first batch of charging stations is as follows: when the number of scheduled charging records is less than a certain threshold (e.g., 500), the system calculates the y (configurable, e.g., 40) nearest charging stations within x km (configurable, e.g., 20 km) of the starting point and obtains the charging station information; when the number is greater than the threshold, the system first sorts the charging stations actually used by the user by the number of times they have been used, selects the z (e.g., 50, configurable) charging stations with the most uses, calculates the distance between the scheduled charging departure point and the selected charging station, and if the distance is less than or equal to the distance threshold (configurable, e.g., 20 km), it is retained as the charging station for the first round of recommendations; if the distance is greater than the distance threshold, it is removed.

[0066] 2. As mentioned above, based on the historical idle data of charging stations, the idle number of fast and slow charging stations and the total idle number of charging stations at the scheduled charging time are predicted. This time point is accurate to the hour. The method used to predict the idle number is time series prediction. The time series model used is not limited to machine learning models such as LSTM / RNN. The model accuracy evaluation standard used is the root mean square error (R 2 ) standard, select the final machine learning model according to the evaluation criteria to predict the number of idle fast and slow charging stations at the appointment time, and select the final first-round recommended charging stations based on the predicted number of idle fast and slow charging stations.

[0067] 3. The pure electric mode power consumption calculation logic described above takes into account the basic battery power consumption, the impact of ambient temperature, and the customer's historical actual power consumption. The basic battery power consumption takes into account the impact of vehicle speed and high-voltage battery SOC. The customer's historical actual power represents the customer's historical actual demand driving power at a specific vehicle speed.

[0068] 4. As described above, based on the map information, there may be multiple paths to the final recommended charging station in the first round. The path information needs to be converted into a speed segment map, and the SOC required to reach the final recommended charging station in the first round needs to be calculated based on the speed segment map and power consumption calculation logic.

[0069] 5. When arriving at the scheduled time as described above, the remaining SOC required to reach the second round of recommended charging stations must be recalculated based on the map's real-time traffic information and the pure electric mode power consumption calculation logic. If the SOC is less than or equal to 0, the customer will be prompted to use the series mode to reach a certain scheduled charging station. If the remaining SOC is greater than 0, the real-time fast and slow charging idle information of the second round of recommended charging stations must be obtained. If it is greater than the fast and slow charging idle number threshold, the remaining SOC is sorted from large to small as the final recommended charging station, and the corresponding driving route is recommended.

[0070] The beneficial effects brought about by the technical solution of the present invention are as follows:

[0071] To address the issue of pre-booked charging for extended-range plug-in hybrid electric vehicles, the present invention designs an intelligent system and method for pre-booked charging for extended-range plug-in hybrid electric vehicles. This system and method consider the vehicle's remaining SOC calculation method, the user's historical driving behavior habits, and the historical and real-time usage information of charging stations, combined with map information, to intelligently recommend the optimal pre-booked charging stations and driving routes for the driver. This system can provide users with the optimal pre-booked and available charging station information and driving route information.

[0072] Definitions of Abbreviations and Key Terms

[0073] VCU: Vehicle Controller;

[0074] PDCU: Power Domain Controller

[0075] VIN: Vehicle Identification Number

[0076] SOC: battery charge

[0077] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

[0078] The contents not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. An intelligent method for pre-charging an extended-range plug-in hybrid electric vehicle, characterized in that: Multiple rounds of recommendations for reserved charging stations include the following steps: First round of recommendations: Select the first batch of recommended charging stations based on the number of scheduled charging records. According to the historical idle data of charging stations, predict the idle fast and slow charging stations and the total idle number of charging stations at the scheduled charging time. And select the first round of final recommended scheduled charging stations based on the idle number threshold. Second round of recommendations: Based on the map information and the power consumption calculation logic in pure electric mode, the remaining SOC at the first round of recommended charging stations is calculated, and the second round of recommended charging stations is recommended based on the remaining SOC; Final recommendation: When the appointment time is reached, the SOC of the charging station recommended for the second round of appointments is calculated based on the real-time traffic information on the map and the power consumption calculation logic of the pure electric mode. The final charging station recommendation is made based on the SOC and the real-time fast and slow charging idle information of the charging station.

2. The intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to claim 1, characterized in that: The specific process of the first round of recommendation steps is as follows: Departure information reservation: input the departure location and departure time of the scheduled charging through the vehicle computer or mobile phone; First, reservation record acquisition: If the number of user reservation charging records in the system is less than n, the system calculates the y nearest charging stations within x km of the starting point as recommended charging stations, and obtains the charging station information; Second, reservation record acquisition: If there are more than n user reservation charging records in the system, first sort the number of times the user has actually used the charging stations, select the z charging stations with the most times of use, calculate the distance between the departure point and the selected charging station, and if it is less than or equal to the distance threshold, keep it as the recommended charging station and obtain the charging station information; if it is greater than the distance threshold, remove it; Information extraction: Extract the number of idle charging guns at the recommended charging station in the past m days from the charging station historical data system.

3. The intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to claim 2, characterized in that: The information extraction step includes the following processes: Data set: The data of the number of idle charging guns for fast and slow charging at recommended charging stations for m days are divided into an input set and a target set. The data of the first m1 days are used as the input data set, and the data of the last m2 days are used as the target data set, where m = m1 + m2, and m1>m2; Filling in empty values ​​of fast and slow charging: If there are empty values ​​in the fast and slow charging data of a station for m1 days, the empty values ​​of fast charging are filled with the average value of idle fast charging data μ1 in m1 days, and the empty values ​​of slow charging data are filled with the average value of idle slow charging data μ2; Normalization: Normalize the fast and slow charging data of days m1 and m2 respectively; Time Series Analysis: Time series machine learning models are used for time series analysis, using the root mean square error R 2 To evaluate time series models, choose R 2 The model with the smallest value predicts the number of fast and slow charging idle stations at a charging station at the driver's reservation time and recommends it as the final charging station in the first round.

4. The intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to claim 3, characterized in that: The normalization step is the min-max method. The normalized fast charge idle number is: (fast – fast min ) / (fast max –fast min ), Similarly, the normalized slow charging idle number is: (slow–slow fast ) / (slow max –slow min ), similarly, use the average value of the fast charge number μ3 and the average value of the slow charge number μ4 of the target set m2 to fill the empty values ​​of the target data set, and use the maximum value of the target set fastTarget max and fastTarget min Denormalize the target set.

5. The intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to claim 4, characterized in that: In the time series analysis step, the process of predicting the number of fast and slow charging idle times at a charging station at the driver's reservation time is as follows: If a station finally predicts that the number of fast-charging idle batteries is > fast0 and the number of slow-charging idle batteries is > slow0, and the total number of idle batteries that is the sum of the number of fast-charging idle batteries and the number of slow-charging idle batteries is > total0, then it is recommended as the final charging station in the first round.

6. The intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to claim 5, characterized in that: In the time series analysis step, the time series machine learning model includes but is not limited to LSTM\RNN model.

7. The intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to claim 6, characterized in that: The specific process of the second round of recommendation steps is as follows: Map information acquisition: Based on the starting point entered by the user and the first round of recommended charging stations, map information from the starting point to each charging station is obtained; Speed ​​segment map conversion: convert the traffic information into a speed segment map based on the actual map traffic information; Remaining SOC calculation: Based on the actual SOC of the vehicle and the pure electric mode driving range calibration table, the actual power consumption can be obtained, and the remaining SOC1 after the vehicle completely passes the first section of the road can be calculated; the remaining SOCn after passing the second section of the road, the third section of the road, and so on can be calculated successively; Pure electric mode selection: If the SOC is less than 0 after a certain section of road, it means that the pure electric mode cannot reach the recommended charging station. Finally, the charging station with the remaining SOCn>0, that is, the charging station that can be reached by the pure electric mode, is selected as the second round of recommended charging stations, and the accessible route information is provided; Series mode selection: If the final SOCn<=0, the user is prompted that the series mode is needed to reach some of the charging stations recommended in the first round that are used as the second round of recommendations.

8. The intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle according to claim 7, characterized in that: The specific process of the final recommended steps is as follows: Pure electric mode SOC calculation: When the user's scheduled time is reached, the actual map information of the route to the second round of recommended charging stations at that time is obtained from the map, and the second round of recommended steps are followed. The remaining SOC when the pure electric mode reaches the second round of recommended charging stations is recalculated in the same way. If the calculated remaining SOC is less than 0, the user is prompted to use the series mode to reach the station. Comparison and sorting recommendation: If the remaining SOC>0, it will be used as a recommended charging station. The fast charging idle number, slow charging idle number and total idle number of the recommended charging station will be queried in the charging station information system. If the fast charging idle number is greater than fastLast, the slow charging idle number is greater than slowLast and the total idle number is greater than totalLast, it can be used as the final recommended charging station. The final recommended charging stations are sorted from large to small according to the remaining SOC as the final recommended charging station list, and the corresponding driving route is recommended.

9. An intelligent device for pre-charging an extended-range plug-in hybrid electric vehicle, storing a computer program, characterized in that: The computer program can execute the intelligent method for scheduled charging of an extended-range plug-in hybrid electric vehicle as described in claims 1 to 8.

10. A computer system, characterized in that: The computer system includes the intelligent device for pre-charging the extended-range plug-in hybrid electric vehicle as claimed in claim 9.

Citation Information

Patent Citations

  • Smart city charging intelligent recommendation method and system

    CN112115350A

  • Automatic driving vehicle intelligent charging method and system based on vehicle-road cooperation

    CN115009073A

  • User charging pile recommendation method and device based on collaborative filtering

    CN115293846A

  • Charging station recommendation method, device and equipment and readable storage medium

    CN115587246A

  • Intelligent reservation charging method and device for extended-range plug-in hybrid electric vehicle and computer system

    CN116562405A

Cited By

  • New energy automobile charging guiding method and system based on controlled Petri network

    CN119737970A

  • Electric vehicle charging station recommendation method and device based on charging preference of driver

    CN120579800A