WO2020253204A1 - 电动车能耗预测方法、计算机可读存储介质和电子设备 - Google Patents

电动车能耗预测方法、计算机可读存储介质和电子设备 Download PDF

Info

Publication number
WO2020253204A1
WO2020253204A1 PCT/CN2019/129473 CN2019129473W WO2020253204A1 WO 2020253204 A1 WO2020253204 A1 WO 2020253204A1 CN 2019129473 W CN2019129473 W CN 2019129473W WO 2020253204 A1 WO2020253204 A1 WO 2020253204A1
Authority
WO
WIPO (PCT)
Prior art keywords
historical
electric vehicle
energy consumption
data
value
Prior art date
Application number
PCT/CN2019/129473
Other languages
English (en)
French (fr)
Inventor
艾建伍
Original Assignee
北京嘀嘀无限科技发展有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京嘀嘀无限科技发展有限公司 filed Critical 北京嘀嘀无限科技发展有限公司
Publication of WO2020253204A1 publication Critical patent/WO2020253204A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/40—Business processes related to the transportation industry

Definitions

  • the invention relates to the technical field of application of new energy vehicles, in particular to a method for predicting energy consumption of electric vehicles, a computer-readable storage medium and electronic equipment.
  • the battery cruising range is a key factor hindering the development of electric vehicles, especially when the winter temperature is low, the cruising range of electric vehicles is greatly reduced. This will cause drivers to "mileage anxiety" and even affect the development and promotion of electric vehicles.
  • a large number of new energy vehicles have been added to the ranks of online car-hailing. If the dispatch platform does not consider the remaining range of new energy vehicles and blindly dispatch orders, it may cause the order mileage to exceed the remaining range of electric vehicles. problem. In this case, the driver can only be forced to cancel the order, which greatly affects the efficiency of the platform and the driver's experience, and further aggravates the driver's "mileage anxiety".
  • electric vehicle energy consumption prediction is very important.
  • the embodiment of the present invention aims to provide an electric vehicle energy consumption prediction method, a computer-readable storage medium, and an electronic device, so as to solve the technical problem that the electric vehicle electricity consumption calculation in the prior art has delay and cannot meet the demand.
  • the present invention provides an electric vehicle energy consumption prediction method, which includes the following steps:
  • each set of the historical trajectory data includes a plurality of position coordinates and a power value corresponding to each of the position coordinates;
  • All the characteristic values of the independent variables and the characteristic values of the dependent variables are correspondingly input into a preset machine learning model to train the machine learning model to obtain an electric energy estimation model for predicting the energy consumption of the electric vehicle.
  • each set of historical trajectory data includes a plurality of position coordinates and the amount of electricity corresponding to each of the position coordinates.
  • the historical travel data of the electric vehicle is obtained in the following manner:
  • each group of the historical trajectory data also includes a sampling time corresponding to the position coordinates and the power value ;
  • the position coordinates include latitude and longitude coordinates and altitude.
  • each set of historical trajectory data includes a plurality of position coordinates and the amount of electricity corresponding to each of the position coordinates.
  • the historical track data is obtained in the following way:
  • Two of the position coordinates are selected as the start point and end point coordinates respectively; the sampling time difference corresponding to the start point and the end point coordinates is less than the set time, or the mileage value between the start point and the end point coordinates is less than the set time.
  • the foregoing method for predicting energy consumption of electric vehicles further includes the following steps:
  • the estimated energy consumption value of the order route is obtained according to the sum of the estimated energy consumption value of each segment of the driving track.
  • the aforementioned method for predicting energy consumption of electric vehicles, in response to order request information, obtaining an order route, and dividing the order route into multiple driving trajectories further includes: obtaining the sending time of the order request;
  • the driving trajectory data includes the mileage value and the altitude difference of the driving trajectory: the driving trajectory data further includes the sending time.
  • the foregoing method for predicting energy consumption of electric vehicles further includes the following steps:
  • each set of historical trajectory data includes a plurality of position coordinates and the amount of electricity corresponding to each of the position coordinates.
  • the step of acquiring historical travel data of online-hailing electric vehicles as the historical travel data of electric vehicles further includes:
  • the historical travel data of the online-hailing electric vehicle is filtered, and the abnormal values therein are eliminated; wherein, if the change in the electric quantity value in the historical travel data of the online-hailing electric vehicle exceeds the normal range value, the historical travel of the online-hailing electric vehicle is determined
  • the data is an abnormal value; or, if the location coordinate or power value is missing in the historical travel data of the online electric vehicle, it is determined that the historical travel data of the online electric vehicle is an abnormal value.
  • the independent variable characteristic value includes a mileage value and/or an altitude difference and/or time.
  • the present invention also provides a computer-readable storage medium in which program instructions are stored, and the computer reads the program instructions and executes any one of the above-mentioned methods for predicting energy consumption of electric vehicles.
  • the present invention also provides an electronic device, including at least one processor and at least one memory, at least one of the memory stores a program instruction, at least one of the processors reads the program instruction and executes any of the above Energy consumption prediction method for electric vehicles.
  • the method includes the following steps: obtaining multiple sets of historical trajectory data according to the historical travel data of the electric vehicle, each set of the historical trajectory data It includes a plurality of position coordinates and a power value corresponding to each of the position coordinates; obtains the characteristic value of the independent variable corresponding to the historical trajectory data according to each of the position coordinate information in each of the historical trajectory data; The power value corresponding to each of the position coordinate data obtains the energy consumption value corresponding to the historical track as the characteristic value of the dependent variable; all the characteristic values of the independent variable and the characteristic values of the dependent variable are correspondingly input to the forecast
  • the machine learning model is trained in the set machine learning model to obtain an electric energy estimation model used to predict the energy consumption of the electric vehicle.
  • a large amount of historical travel data of electric vehicles can be used, and training sample data can be extracted therefrom to train the machine learning model to obtain an electric energy estimation model.
  • Using the electric energy prediction model to predict the energy consumption of electric vehicles can predict the electric energy consumption based on the vehicle driving data in advance, avoiding the problems caused by the delay of the prediction results.
  • FIG. 1 is a flowchart of a method for predicting energy consumption of electric vehicles according to an embodiment of the present invention, which mainly shows the modeling process of energy consumption prediction of electric vehicles;
  • FIG. 2 is a flowchart of a method for predicting energy consumption of electric vehicles according to an embodiment of the present invention
  • FIG. 3 is a schematic diagram of splitting orders and estimating order power consumption according to an embodiment of the present invention.
  • FIG. 4 is a schematic diagram of the hardware connection relationship of the electronic device according to an embodiment of the present invention.
  • This embodiment provides an electric vehicle energy consumption prediction method, which can be applied to a cloud server, which can realize data communication with the electric vehicle driving database, and can also realize data communication with the on-board control device on the electric vehicle, that is, the cloud server can All the driving data of related electric vehicles can be obtained, and the current state data of each electric vehicle can also be obtained.
  • the method includes the following steps:
  • each set of the historical trajectory data includes multiple position coordinates and a power value corresponding to each of the position coordinates; as described above, the cloud server can acquire
  • the historical data of each electric vehicle during the driving process includes at least the electric vehicle driving route, its power consumption, driving speed, time, etc.
  • the historical travel data of the electric vehicle can be obtained in the following manner: during the driving of the electric vehicle, the position coordinates and the power value are collected according to a preset sampling time (the power value can be the SOC of the battery).
  • Each group of the historical trajectory data also includes a sampling time corresponding to the position coordinates and the power value; the position coordinates include latitude and longitude coordinates and altitude.
  • the preset sampling time can be about ten seconds.
  • the longitude, latitude and altitude can be directly collected by a positioning sensor installed on the vehicle. By considering the altitude, the impact of the climbing process and the downhill process of the vehicle on the power consumption during the driving process can also be integrated into the power consumption prediction model, so that the model is more suitable for the actual situation.
  • S102 Obtain the characteristic value of the independent variable corresponding to the historical trajectory data according to each of the position coordinate information in each of the historical trajectory data; obtain the characteristic value of the independent variable corresponding to each of the position coordinate data;
  • the energy consumption value corresponding to the historical trajectory is used as the characteristic value of the dependent variable; through the analysis of the historical trajectory data, it is possible to determine the characteristics closely related to the energy consumption of the electric vehicle, such as mileage, time, etc., and then use it as the characteristic value of the independent variable.
  • the power consumption value corresponding to the characteristic value of the variable can be used as the characteristic value of the dependent variable.
  • S103 Correspondingly input all the characteristic values of the independent variables and the characteristic values of the dependent variables into a preset machine learning model to train the machine learning model to obtain an electric energy estimate used to predict the energy consumption of electric vehicles model.
  • the driving data records of electric vehicles are acquired every interval, such as every three seconds or every ten seconds. Therefore, the amount of data that can be included in the historical trip data is very large, that is, It is said that when the machine learning model is trained, the amount of training sample models is very large, which can ensure the accuracy of machine learning model training.
  • a suitable machine learning model can be selected according to the type and number of independent variables in historical driving data.
  • the Wide&Deep model can be selected.
  • the Wide&Deep model belongs to a deep machine learning model widely used in the prior art.
  • the Wide branch performs second-order crossover of features and has a certain memory function for historical data
  • the Deep branch is The traditional multi-layer perceptron structure has good generalization ability.
  • the two branches are used in combination, and they can learn from each other. After the machine learning model is selected, the input and output variables of the machine learning model can be determined.
  • the independent variables selected in step S102 are used as the input of the machine learning model, and the dependent variables selected in step S102 are used as the machine learning model.
  • the machine learning model can train its internal parameter values after inputting a large amount of sample data. Since the machine learning process belongs to the prior art, it will not be described in detail in this step. After a large amount of iterative training, the machine learning model is adapted to the electric energy consumption prediction of the electric vehicle.
  • a large amount of historical travel data of electric vehicles can be used to train the machine learning model to obtain an electric energy estimation model.
  • Using the electric energy prediction model to predict the energy consumption of electric vehicles can predict the electric energy consumption based on the vehicle driving data in advance, avoiding the problems caused by the delay of the prediction results.
  • the historical travel data of the online-hailing electric vehicle is acquired as the historical travel data of the electric vehicle.
  • a large amount of online car-hailing electric vehicle driving data can be obtained in the online car-hailing dispatching platform.
  • Each net-hailing electric vehicle is equipped with a mobile terminal for real-time communication with the net-hailing dispatching platform, so the net-hailing dispatching platform can obtain the actual driving data of each net-hailing electric vehicle in real time.
  • Using these data directly as the historical travel data of the electric vehicle in step S101 is faster and simplifies the data acquisition process.
  • the method may further include the following steps: filtering the historical travel data of the online-hailing electric vehicle to eliminate abnormal values; wherein, if the change in the electric quantity value in the historical travel data of the online-hailing electric vehicle exceeds the normal range value It is determined that the historical itinerary data of the online electric car is an abnormal value; if the position coordinate or the power value is missing in the historical itinerary data of the online electric car, it is determined that the historical itinerary data of the online electric car is an abnormal value.
  • filtering the historical travel data of the online-hailing electric vehicle to eliminate abnormal values
  • the change in the electric quantity value in the historical travel data of the online-hailing electric vehicle exceeds the normal range value
  • the position coordinate or the power value is missing in the historical itinerary data of the online electric car
  • the historical itinerary data of the online electric car is an abnormal value.
  • there is a specific relationship between its driving path, time and power consumption value For example, the longer the driving path of the electric car-hailing,
  • the historical trajectory data in step S101 can be obtained through the following steps:
  • S201 Select two of the position coordinates as the start point coordinate and the end point coordinate respectively; the sampling time difference corresponding to the start point coordinate and the end point coordinate is less than the set time, or the mileage value between the start point coordinate and the end point coordinate Less than the set mileage; this step is mainly to select the start and end points of the historical trajectory, and use the historical trajectory with a smaller mileage as the standard data for calculation as much as possible, so that when predicting the power consumption of a certain planned mileage in the future, The planned mileage can be divided into multiple segments that are close to the historical trajectory in order to realize the estimation of electric energy consumption. Therefore, the above set time can be selected as a few minutes or half an hour, etc., and the set mileage can be selected from several kilometers to tens of kilometers.
  • S202 Determine the historical track between the start point coordinates and the end point coordinates and all the position coordinates covered by the historical track; because there can be many routes from the start point to the end point, only the coordinates between the two The location is determined to get the true historical track. Therefore, in order to uniquely determine the historical trajectory, each coordinate position covered by the historical trajectory needs to be determined.
  • S203 Obtain the mileage value and the altitude difference of the historical trajectory according to all the position coordinates covered by the historical trajectory as the historical trajectory data.
  • the mileage value can be obtained by summing the distance between every two coordinate positions in the historical trajectory
  • the altitude difference can be obtained by summing the altitude difference between no two coordinate positions in the historical trajectory.
  • the mileage value and altitude difference can be used as the characteristic value of the independent variable of a piece of historical trajectory; the difference between the power value corresponding to the starting point coordinate and the power value corresponding to the ending point coordinate in the historical trajectory is used as the characteristic value of the dependent variable .
  • Using the above model training method is simple and easy to implement.
  • the above method further includes the step of using an electric energy estimation model to predict the electric energy that may be consumed by a certain order, specifically including:
  • S104 In response to the order request information, obtain the order route, and divide the order route into multiple driving trajectories; the order request information is sent by passengers according to their actual needs, which can be realized by using APP installed on terminals such as mobile phones.
  • the order route can be divided according to the historical trajectory selection principle, that is, the travel trajectory obtained after the division meets its required travel time less than the set time or its mileage value is less than the set mileage value. Because the data related to the divided driving trajectory will finally be input to the electric energy estimation model as an independent variable, it is ensured that the data of the divided driving trajectory has a high degree of conformity with the historical trajectory data when the electric energy estimation model is trained.
  • S105 Acquire driving trajectory data of each segment of the driving trajectory, where the driving trajectory data includes the mileage value and altitude difference of the driving trajectory; when the order route is determined, the entire route can be displayed on the electronic map Therefore, the latitude, longitude and altitude of each position coordinate in the order route can also be obtained, so this step can be implemented directly using the navigation device, electronic map or other positioning sensors on the vehicle.
  • S106 Obtain the estimated energy consumption value of each segment of the driving trajectory according to the driving trajectory data of each segment of the driving trajectory and the electric energy estimation model; substitute each driving trajectory data in the order route into the electric energy estimation model , It is possible to determine the amount of change in the SOC value required for this segment of the trajectory.
  • S107 Obtain the estimated energy consumption value of the order route according to the sum of the estimated energy consumption value of each segment of the driving track.
  • the SOC value changes of all the driving trajectories in the order route are summed to obtain the estimated energy consumption value corresponding to the order route.
  • the above technical solution can quickly estimate the energy consumption value corresponding to the order request after confirming the order request, that is, when the order request is received, the power consumption of the order can be determined, there is no delay, and this step is adopted
  • the method in can greatly reduce the amount of data calculations and improve the efficiency of predicting order power consumption.
  • the solution in this embodiment divides the data in each historical itinerary into multiple small historical trajectories.
  • the electric vehicle driving situation of each historical trajectory can be used as a standardized module.
  • the path in the new order can be split according to the historical trajectory in the standardized module, if it can be standardized with the historical trajectory
  • the modules overlap when the starting point and end point of the driving trajectory are input, the electric energy consumption estimation result of this segment of the driving trajectory can be directly obtained.
  • the starting point and end point of the driving trajectory are input, it is determined that it belongs to a new path, and the machine can Learn the model to calculate the power consumption of the new driving trajectory.
  • the order route A1-A102 by splitting the order route, except for the two driving trajectories A1-A2 and A101-A102, the rest of the driving trajectories can be compared with the existing history If the trajectory coincides, when calculating the power consumption, you only need to recalculate the power consumption of the two driving trajectories A1-A2 and A101-A102.
  • the power consumption of the remaining driving trajectories can be directly derived from the power consumption corresponding to the historical trajectory .
  • the above scheme may also include the following steps:
  • S108 Send the estimated energy consumption value corresponding to the order route to the order dispatch platform.
  • the electricity consumption of orders predicted by this scheme is more immediate and there is no problem of delay. Therefore, the forecast results are sent to the dispatch platform for reference by the dispatch platform, and the dispatch platform can assist the dispatch platform to select vehicles, which can improve the efficiency and accuracy of dispatch.
  • step S104 further includes acquiring the sending time of the order request; in the step S105, the driving track data further includes the sending time.
  • time is used as a characteristic value of an independent variable to train the machine learning model, so that battery consumption can be predicted according to different seasons.
  • the data for the summer and winter seasons can be grouped. Because the temperature is lower in winter, the remaining cruising range that can be driven by the same battery power will be reduced. Therefore, the historical itinerary data is grouped according to winter and summer.
  • a new order request is obtained, it is first judged whether the order request corresponds to summer time or winter time. Combined with the sending time of the order request, it is substituted into the energy estimation model, and the time As a reference factor to estimate power consumption, more accurate analysis results can be obtained.
  • This embodiment provides a computer-readable storage medium in which a computer program is stored, and the computer program is executed by a computer to implement the electric vehicle energy consumption prediction method described in any one of the technical solutions in Embodiment 1.
  • This embodiment provides an electronic device. As shown in FIG. 4, it includes at least one processor 401 and at least one memory 402. At least one of the memories 402 stores instruction information, and at least one of the processors 401 reads the After the program instructions, the electric vehicle energy consumption prediction method described in any one of Embodiments 1 or 2 can be executed.
  • the above-mentioned device may further include: an input device 403 and an output device 404.
  • the processor 401, the memory 402, the input device 403, and the output device 404 may be connected by a bus or other methods.
  • the above-mentioned products can execute the methods provided in the embodiments of the present application, and have functional modules and beneficial effects corresponding to the execution methods. For technical details not described in detail in this embodiment, please refer to the method provided in the embodiment of this application.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Evolutionary Computation (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Operations Research (AREA)
  • Medical Informatics (AREA)
  • Quality & Reliability (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Electric Propulsion And Braking For Vehicles (AREA)
  • Navigation (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种电动车能耗预测方法、计算机可读存储介质和电子设备，其中的方法包括如下步骤：根据电动车历史行程数据获取多组历史轨迹数据（S101）；根据每一历史轨迹数据中的每一位置坐标信息得到与历史轨迹数据对应的自变量特征值；根据与每一位置坐标数据对应的电量值得到与历史轨迹对应的能耗值作为因变量特征值（S102）；将自变量特征值和因变量特征值对应地输入至预设的机器学习模型中对机器学习模型进行训练，得到用于预测电动车能耗的电能预估模型（S103）。以上方案，根据电动车历史行程数据提取出的特征对机器学习模型进行训练从而得到电能预估模型，采用电能预估模型对电动车能耗进行预测能根据车辆行驶数据预先对电能消耗进行预测，避免预测结果延时。

Description

电动车能耗预测方法、计算机可读存储介质和电子设备
本申请要求在2019年06月17日提交中国专利局、申请号为201910521895.6、发明名称为“电动车能耗预测方法、计算机可读存储介质和电子设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及新能源车辆应用技术领域，具体涉及一种电动车能耗预测方法、计算机可读存储介质和电子设备。
背景技术
目前，电池续航里程是阻碍电动汽车发展的关键因素，尤其是冬季气温较低时，电动车的续航里程更是大打折扣。这会导致司机“里程焦虑”，甚至影响电动汽车的发展和推广。另一方面，目前已经有大量的新能源车加入到网约车行列中，如果派单平台不考虑新能源车的剩余续航里程就盲目派单，可能会造成订单里程超过电动汽车剩余续航里程的问题。这种情况下，司机也只能被迫取消订单，极大影响了平台的派单效率和司乘体验，而且进一步加剧了司机的“里程焦虑”。综上，对于司机、网约车派单平台以及汽车厂商来说，电动车能耗预测都是非常重要的事情。
现有技术中在获得电动车剩余续航里程时都要考虑电池本身的属性特征，比如电压、电流和内阻等，但是这类数据很难实时获取，其获取到这类数据后计算得到的续航里程具有一定的延迟，所以无法真正满足网约车派单平台或者司机的需求。
发明内容
本发明实施例旨在提供一种电动车能耗预测方法、计算机可读存储介质和电子设备，以解决现有技术中电动汽车电量消耗计算具有延 迟性无法满足需求的技术问题。
为此，本发明提供一种电动车能耗预测方法，包括如下步骤：
根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值；
根据每一所述历史轨迹数据中的每一所述位置坐标信息得到与所述历史轨迹数据对应的自变量特征值；根据与每一所述位置坐标数据对应的电量值得到与所述历史轨迹对应的能耗值作为因变量特征值；
将所有的所述自变量特征值和所述因变量特征值对应地输入至预设的机器学习模型中对所述机器学习模型进行训练，得到用于预测电动车能耗的电能预估模型。
可选地，上述的电动车能耗预测方法，根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值的步骤中，所述电动车历史行程数据通过如下方式得到：
在所述电动车行驶过程中，按照预设采样时间采集所述位置坐标和所述电量值；每一组所述历史轨迹数据中还包括与所述位置坐标和所述电量值对应的采样时间；所述位置坐标包括经纬度坐标和海拔高度。
可选地，上述的电动车能耗预测方法，根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值的步骤中，所述历史轨迹数据通过如下方式得到：
选择两个所述位置坐标分别作为起点坐标和终点坐标；所述起点坐标和所述终点坐标对应的采样时间差小于设定时长，或者所述起点坐标与所述终点坐标之间的里程值小于设定里程；
确定所述起点坐标和所述终点坐标之间的历史轨迹以及所述历史轨迹覆盖的所有位置坐标；
根据所述历史轨迹覆盖的所有位置坐标得到所述历史轨迹的里 程值和海拔高度差作为所述历史轨迹数据。
可选地，上述的电动车能耗预测方法，还包括如下步骤：
响应于订单请求信息，获取订单路线，将所述订单路线划分为多段行驶轨迹；
获取每一段所述行驶轨迹的行驶轨迹数据，所述行驶轨迹数据包括所述行驶轨迹的里程值和海拔高度差；
根据每一段所述行驶轨迹的行驶轨迹数据和所述电能预估模型得到每一段所述行驶轨迹的预估能耗值；
根据每一段所述行驶轨迹的预估能耗值的和得到所述订单路线的预估能耗值。
可选地，上述的电动车能耗预测方法，响应于订单请求信息，获取订单路线，将所述订单路线划分为多段行驶轨迹的步骤中还包括：获取订单请求的发送时间；
获取每一段所述行驶轨迹的行驶轨迹数据，所述行驶轨迹数据包括所述行驶轨迹的里程值和海拔高度差的步骤中：所述行驶轨迹数据还包括所述发送时间。
可选地，上述的电动车能耗预测方法，还包括如下步骤：
发送所述订单路线的预估能耗值至派单平台。
可选地，上述的电动车能耗预测方法，根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值的步骤中：
获取网约电动车的历史行程数据作为所述电动车历史行程数据。
可选地，上述的电动车能耗预测方法，获取网约电动车的历史行程数据作为所述电动车历史行程数据的步骤中还包括：
对所述网约电动车历史行程数据进行筛选，剔除其中的异常值；其中，若所述网约电动车历史行程数据中的电量值变化超出正常范围值则判定所述网约电动车历史行程数据为异常值；或者，若所述网约电动车历史行程数据中缺失位置坐标或者电量值则判定所述网约电动车历史行程数据为异常值。
可选地，上述的电动车能耗预测方法，所述自变量特征值包括里 程值和/或海拔高度差和/或时间。
本发明还提供一种计算机可读存储介质，所述存储介质中存储有程序指令，计算机读取所述程序指令后执行以上任一项所述的电动车能耗预测方法。
本发明还提供一种电子设备，包括至少一个处理器和至少一个存储器，至少一个所述存储器中存储有程序指令，至少一个所述处理器读取所述程序指令后执行以上任一项所述的电动车能耗预测方法。
与现有技术相比，本发明实施例提供的上述技术方案至少具有以下有益效果：
本发明实施例提供的电动车能耗预测方法、计算机可读存储介质和电子设备，其中的方法包括如下步骤：根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值；根据每一所述历史轨迹数据中的每一所述位置坐标信息得到与所述历史轨迹数据对应的自变量特征值；根据与每一所述位置坐标数据对应的电量值得到与所述历史轨迹对应的能耗值作为因变量特征值；将所有的所述自变量特征值和所述因变量特征值对应地输入至预设的机器学习模型中对所述机器学习模型进行训练，得到用于预测电动车能耗的电能预估模型。采用本发明提供的以上方案，能够利用大量的电动车历史行程数据，从中提取出训练样本数据对机器学习模型进行训练从而得到电能预估模型。采用电能预估模型对电动车能耗进行预测能够预先根据车辆行驶数据对电能消耗进行预测，避免预测结果延时所产生的问题。
附图说明
图1为本发明一个实施例所述电动车能耗预测方法的流程图，主要表明了电动车能耗预测的建模过程；
图2为本发明一个实施例所述对电动车能耗进行预测的方法流程图；
图3为本发明一个实施例所述对订单进行拆分并预估订单电量消耗的原理图；
图4为本发明一个实施例所述电子设备的硬件连接关系示意图。
具体实施方式
下面将结合附图进一步说明本发明实施例。在本发明的描述中，需要说明的是，术语“中心”、“上”、“下”、“左”、“右”、“竖直”、“水平”、“内”、“外”等指示的方位或位置关系为基于附图所示的方位或位置关系，仅是为了便于描述本发明的简化描述，而不是指示或暗示所指的装置或组件必需具有特定的方位、以特定的方位构造和操作，因此不能理解为对本发明的限制。此外，术语“第一”、“第二”、“第三”仅用于描述目的，而不能理解为指示或暗示相对重要性。其中，术语“第一位置”和“第二位置”为两个不同的位置。
在本发明的描述中，需要说明的是，除非另有明确的规定和限定，术语“安装”、“相连”、“连接”应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或一体地连接；可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连，可以是两个组件内部的连通。对于本领域的普通技术人员而言，可以具体情况理解上述术语在本发明中的具体含义。
本发明提供的以下实施例中的各个技术方案，除非彼此之间相互矛盾，否则不同技术方案之间可以相互组合，其中的技术特征可以相互替换。
实施例1
本实施例提供一种电动车能耗预测方法，可应用于云端服务器，该服务器能够与电动车行驶数据库实现数据通信，也能够与电动车上的车载控制器件实现数据通信，即该云端服务器可以获取到相关电动车的所有行驶数据，也能够获取到每一电动车的当前状态数据。具体地，如图1所示，该方法包括如下步骤：
S101：根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值；如前所述，云端服务器能够获取每一电动车行驶过程中的历史数据，其中至少包括电动车行驶路线、其消耗电能、行驶速度、 时间等。具体地，可以通过如下方式得到所述电动车历史行程数据：在所述电动车行驶过程中，按照预设采样时间采集所述位置坐标和所述电量值(所述电量值可以采用电池的SOC值来表示)；每一组所述历史轨迹数据中还包括与所述位置坐标和所述电量值对应的采样时间；所述位置坐标包括经纬度坐标和海拔高度。所述预设采样时间可以选择十秒左右。所述经纬度和海拔高度可以通过设置于车辆上的定位传感器直接采集得到。通过考虑海拔高度能将车辆行驶过程中的爬坡过程和下坡过程对电能消耗的影响也结合到电能消耗预估模型中，从而使模型与实际情况更加贴合。
S102：根据每一所述历史轨迹数据中的每一所述位置坐标信息得到与所述历史轨迹数据对应的自变量特征值；根据与每一所述位置坐标数据对应的电量值得到与所述历史轨迹对应的能耗值作为因变量特征值；通过对历史轨迹数据进行解析，能够确定和电动车能耗关联紧密的特征如行驶里程、时间等，之后将其作为自变量特征值，与自变量特征值对应的电量消耗值即可作为因变量特征值。
S103：将所有的所述自变量特征值和所述因变量特征值对应地输入至预设的机器学习模型中对所述机器学习模型进行训练，得到用于预测电动车能耗的电能预估模型。如前所述，对于电动车的行驶数据记录是每间隔一个周期获取一次的，例如每三秒钟或者每十秒钟获取一次等，因此历史行程数据中可以包括的数据量非常大，也就是说对机器学习模型进行训练时的训练样本模型量很大，能够确保机器学习模型训练的准确度。
现有技术中，机器学习模型的种类很多，在选择时可以根据历史行驶数据中自变量的种类、数量等来选择适合的机器学习模型。本实施例中可选择Wide&Deep模型，Wide&Deep模型属于现有技术中广泛应用的一种深度机器学习模型，其中的Wide分支对特征进行二阶交叉，对历史数据拥有一定的记忆功能，而Deep分支就是传统的多层感知机结构，有较好的泛化能力，两个分支组合利用，能互取所长。当选定机器学习模型之后，就可以确定好机器学习模型的输入和输出变量，采用步骤S102中选定的自变量作为机器学习模型的输入，步 骤S102中选定的因变量作为机器学习模型的输出即可，机器学习模型能够在大量的样本数据输入后自行对其内部的参数值进行训练。由于机器学习的过程属于现有技术，本步骤中不再详细叙述，经过大量的迭代训练确保机器学习模型适应于电动车的电能消耗预测。
采用本实施例提供的以上方案，能够利用大量的电动车历史行程数据，对机器学习模型进行训练从而得到电能预估模型。采用电能预估模型对电动车能耗进行预测能够预先根据车辆行驶数据对电能消耗进行预测，避免预测结果延时所产生的问题。
优选地，在以上方案中，步骤S101中，获取网约电动车的历史行程数据作为所述电动车历史行程数据。在网约车调度平台中能够得到大量的网约电动车行驶数据。每一台网约电动车上都配置有移动终端与网约车调度平台进行实时通信，因此网约车调度平台可以实时获取到每一网约电动车的实际行驶数据。采用这些数据直接作为步骤S101中的电动车历史行程数据更加快捷，简化了数据获取流程。
优选地，还可以包括如下步骤：对所述网约电动车历史行程数据进行筛选，剔除其中的异常值；其中，若所述网约电动车历史行程数据中的电量值变化超出正常范围值则判定所述网约电动车历史行程数据为异常值；若所述网约电动车历史行程数据中缺失位置坐标或者电量值则判定所述网约电动车历史行程数据为异常值。对于每一电动网约车来说，其行驶路径、时间和电量消耗值之间存在特定的关系，例如当电动网约车的行驶路径越长时其对应的电量消耗值必然是越大的，如果某一历史订单数据中对于同一电动网约车来说，其行驶路径对电量消耗值的影响关系消失或者明显违背常规，则其一定是异常数据。而本实施例中的上述技术方案，需要利用历史轨迹数据中的每一个位置坐标以及对应的电量值、采集时间等数据相结合才能对机器学习模型训练，因此如果以上数据中有所缺失，则可以认为其为异常值。基于同样的原理，还可以采用其他车辆使用规律对异常值进行筛选并删除。删除异常值之后，能够确保得到的样本数据更符合实际情况，训练得到的模型也更加准确。
上述方案中，所述步骤S101中的所述历史轨迹数据可以通过如 下步骤得到：
S201：选择两个所述位置坐标分别作为起点坐标和终点坐标；所述起点坐标和所述终点坐标对应的采样时间差小于设定时长，或者所述起点坐标与所述终点坐标之间的里程值小于设定里程；本步骤中主要是选定历史轨迹的起点和终点，尽可能地采用里程值较小的历史轨迹作为标准数据进行计算，如此将来预测某一计划行驶里程的耗电量时，可以将计划行驶里程拆分成多段和历史轨迹接近的行驶轨迹，以便实现电能消耗的预估。因此，以上设定时长可以选为几分钟或者半小时等，所述设定里程可以选为几公里到几十公里不等。
S202：确定所述起点坐标和所述终点坐标之间的历史轨迹以及所述历史轨迹覆盖的所有位置坐标；因为从起点到终点的路线可以有很多条，只有将二者之间的每一个坐标位置都确定下来才能得到真正的历史轨迹。因此，为了唯一确定历史轨迹，需要确定历史轨迹所覆盖的每一个坐标位置。
S203：根据所述历史轨迹覆盖的所有位置坐标得到所述历史轨迹的里程值和海拔高度差作为所述历史轨迹数据。里程值可以根据历史轨迹中每两个坐标位置之间的距离求和得到，海拔高度差可以根据历史轨迹中没两个坐标位置之间的海拔高度差求和之后得到。
也即，可以采用里程值、海拔高度差作为一段历史轨迹的自变量特征值；采用这段历史轨迹中起点坐标对应的电量值、终点坐标对应的电量值之间的差值作为因变量特征值。针对所有的历史轨迹均做上述特征分解，之后将自变量特征值和因变量特征值输入至机器学习模型中训练即可。采用以上模型训练方法，简单容易实现。
优选地，如图2所示，以上方法还包括利用电能预估模型来预测某一订单可能消耗的电量的步骤，具体地包括：
S104：响应于订单请求信息，获取订单路线，将所述订单路线划分为多段行驶轨迹；订单请求信息为乘客根据自己的实际需求发送，其采用手机等终端上安装的APP即可实现。而订单路线的划分可以根据历史轨迹的选取原则进行划分，也即划分后得到的行驶轨迹满足其所需要的行驶时间小于设定时间或者其里程值小于设定里程值。因为 划分后的行驶轨迹的相关数据最后会作为自变量输入至电能预估模型，因此确保划分后的行驶轨迹的数据与训练所述电能预估模型时的历史轨迹数据具有较高的契合度。
S105：获取每一段所述行驶轨迹的行驶轨迹数据，所述行驶轨迹数据包括所述行驶轨迹的里程值和海拔高度差；当订单路线确定后，就可以在电子地图上将整条路线显示出来，因此也能够得到订单路线中每一个位置坐标的经纬度和海拔高度，所以本步骤可以直接利用车辆上的导航装置、电子地图或者其他定位传感器实现即可。
S106：根据每一段所述行驶轨迹的行驶轨迹数据和所述电能预估模型得到每一段所述行驶轨迹的预估能耗值；将订单路线中的每一行驶轨迹数据代入至电能预估模型中，就能够确定该段行驶轨迹所需要SOC值变化量。
S107：根据每一段所述行驶轨迹的预估能耗值的和得到所述订单路线的预估能耗值。将所述订单路线中的所有所述行驶轨迹的SOC值变化量取和得到与所述订单路线对应的预估能耗值。
采用以上技术方案能够在确定订单请求后迅速预估出与订单请求对应的能耗值，也即在收到订单请求时就能确定订单需要消耗的电能，不存在延时性，而且采用本步骤中的方法，能够极大地降低数据运算量，提高对订单电能消耗预测的效率。
另外，以上方案还可以变换为如下方式：
由于每一订单其对应的起点和终点都是随机的，不一定和历史行程中的数据完全重合，因此本实施例中的方案将每一历史行程中的数据划分为多个小的历史轨迹，每一段历史轨迹的电动车行驶情况可以作为一个标准化模块，当用户发来新的订单时，可以将新的订单中的路径按照标准化模块中的历史轨迹进行拆分，如果能够和历史轨迹的标准化模块重合，则当输入行驶轨迹的起点和终点后就能够直接得出该段行驶轨迹的电能消耗预估结果，当输入行驶轨迹的起点和终点后确定其属于一段新的路径，则可以利用机器学习模型来计算该新的行驶轨迹的电能消耗。如图3所示，其中订单路线A1-A102，通过将订单路线进行拆分，除了其中的A1-A2和A101-A102这两段行驶轨迹之 外，其余的行驶轨迹都可以和已有的历史轨迹重合，则在计算电能消耗时，只需要重新计算A1-A2和A101-A102这两段行驶轨迹的电能消耗就可以了，其余行驶轨迹的电能消耗能够根据历史轨迹对应的电能消耗直接得出。
优选地，以上方案中还可以包括如下步骤：
S108：发送所述与所述订单路线对应的预估能耗值至派单平台。
通过本方案预测的订单电能消耗更加即时，不存在延时的问题，因此将预测结果发送至派单平台供派单平台参考，辅助派单平台选择车辆，能够提高派单效率和准确度。
进一步地，在以上方案中，在步骤S104中还包括，获取订单请求的发送时间；在所述步骤S105中，所述行驶轨迹数据还包括所述发送时间。
采用本方案，将时间作为一个自变量特征值来对机器学习模型进行训练，从而能够根据不同的季节来预测电池消耗量。例如，可以将夏季和冬季两个季节的数据进行分组，因为冬季时节气温较低，同样的电池剩余电量能够行驶的剩余续航里程会有所降低。因此，历史行程数据按照冬季和夏季分组，当获取到新的订单请求时，也首先判断订单请求是对应于夏季时间还是冬季时间，结合订单请求的发送时间代入到电能预估模型中，将时间作为一个参考因素来预估电量消耗，能得到更准确的分析结果。
实施例2
本实施例提供一种计算机可读存储介质，所述存储介质中存储有计算机程序，所述计算机程序被计算机执行后实现实施例1中任一技术方案所述的电动车能耗预测方法。
实施例3
本实施例提供一种电子设备，如图4所示，包括至少一个处理器401和至少一个存储器402，至少一个所述存储器402中存储有指令信息，至少一个所述处理器401读取所述程序指令后可执行实施例1或2中任一方案所述的电动车能耗预测方法。
上述装置还可以包括：输入装置403和输出装置404。处理器401、 存储器402、输入装置403和输出装置404可以通过总线或者其他方式连接。上述产品可执行本申请实施例所提供的方法，具备执行方法相应的功能模块和有益效果。未在本实施例中详尽描述的技术细节，可参见本申请实施例所提供的方法。
最后应说明的是：以上实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims (11)

  1. 一种电动车能耗预测方法，其特征在于，包括如下步骤：
    根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值；
    根据每一组所述历史轨迹数据中的每一所述位置坐标信息得到与所述历史轨迹数据对应的自变量特征值；根据与每一所述位置坐标数据对应的电量值得到与所述历史轨迹对应的能耗值作为因变量特征值；
    将所有的所述自变量特征值和所述因变量特征值对应地输入至预设的机器学习模型中对所述机器学习模型进行训练，得到用于预测电动车能耗的电能预估模型。
  2. 根据权利要求1所述的电动车能耗预测方法，其特征在于，根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值的步骤中，所述电动车历史行程数据通过如下方式得到：
    在所述电动车行驶过程中，按照预设采样时间采集所述位置坐标和所述电量值；每一组所述历史轨迹数据中还包括与所述位置坐标和所述电量值对应的采样时间；所述位置坐标包括经纬度坐标和海拔高度。
  3. 根据权利要求2所述的电动车能耗预测方法，其特征在于，根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值的步骤中，所述历史轨迹数据通过如下方式得到：
    选择两个所述位置坐标分别作为起点坐标和终点坐标；所述起点坐标和所述终点坐标对应的采样时间差小于设定时长，或者所述起点坐标与所述终点坐标之间的里程值小于设定里程；
    确定所述起点坐标和所述终点坐标之间的历史轨迹以及所述历史轨迹覆盖的所有位置坐标；
    根据所述历史轨迹覆盖的所有位置坐标得到所述历史轨迹的里程值和海拔高度差作为所述历史轨迹数据。
  4. 根据权利要求3所述的电动车能耗预测方法，其特征在于，还包括如下步骤：
    响应于订单请求信息，获取订单路线，将所述订单路线划分为多段行驶轨迹；
    获取每一段所述行驶轨迹的行驶轨迹数据，所述行驶轨迹数据包括所述行驶轨迹的里程值和海拔高度差；
    根据每一段所述行驶轨迹的行驶轨迹数据和所述电能预估模型得到每一段所述行驶轨迹的预估能耗值；
    根据每一段所述行驶轨迹的预估能耗值的和得到所述订单路线的预估能耗值。
  5. 根据权利要求4所述的电动车能耗预测方法，其特征在于：
    响应于订单请求信息，获取订单路线，将所述订单路线划分为多段行驶轨迹的步骤中还包括：获取订单请求的发送时间；
    获取每一段所述行驶轨迹的行驶轨迹数据，所述行驶轨迹数据包括所述行驶轨迹的里程值和海拔高度差的步骤中：所述行驶轨迹数据还包括所述发送时间。
  6. 根据权利要求4或5所述的电动车能耗预测方法，其特征在于，还包括如下步骤：
    发送所述订单路线的预估能耗值至派单平台。
  7. 根据权利要求6所述的电动车能耗预测方法，其特征在于，根据电动车历史行程数据获取多组历史轨迹数据，每一组所述历史轨迹数据包括多个位置坐标以及与每一所述位置坐标对应的电量值的步骤中：
    获取网约电动车的历史行程数据作为所述电动车历史行程数据。
  8. 根据权利要求7所述的电动车能耗预测方法，其特征在于，获取网约电动车的历史行程数据作为所述电动车历史行程数据的步骤中还包括：
    对所述网约电动车历史行程数据进行筛选，剔除其中的异常值； 其中，若所述网约电动车历史行程数据中的电量值变化超出正常范围值则判定所述网约电动车历史行程数据为异常值；或者，若所述网约电动车历史行程数据中缺失位置坐标或者电量值则判定所述网约电动车历史行程数据为异常值。
  9. 根据权利要求1所述的电动车能耗预测方法，其特征在于，所述自变量特征值包括里程值和/或海拔高度差和/或时间。
  10. 一种计算机可读存储介质，其特征在于，所述存储介质中存储有程序指令，计算机读取所述程序指令后执行权利要求1-9任一项所述的电动车能耗预测方法。
  11. 一种电子设备，其特征在于，包括至少一个处理器和至少一个存储器，至少一个所述存储器中存储有程序指令，至少一个所述处理器读取所述程序指令后执行权利要求1-9任一项所述的电动车能耗预测方法。
PCT/CN2019/129473 2019-06-17 2019-12-27 电动车能耗预测方法、计算机可读存储介质和电子设备 WO2020253204A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910521895.6 2019-06-17
CN201910521895.6A CN110222906A (zh) 2019-06-17 2019-06-17 电动车能耗预测方法、计算机可读存储介质和电子设备

Publications (1)

Publication Number Publication Date
WO2020253204A1 true WO2020253204A1 (zh) 2020-12-24

Family

ID=67817422

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/129473 WO2020253204A1 (zh) 2019-06-17 2019-12-27 电动车能耗预测方法、计算机可读存储介质和电子设备

Country Status (2)

Country Link
CN (1) CN110222906A (zh)
WO (1) WO2020253204A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115061462A (zh) * 2022-05-18 2022-09-16 安徽亚珠金刚石股份有限公司 一种人造金刚石车间用agv车体自动充电规划系统
CN115808922A (zh) * 2022-01-07 2023-03-17 宁德时代新能源科技股份有限公司 商用电动车辆能耗预测方法、装置和计算机设备

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110222906A (zh) * 2019-06-17 2019-09-10 北京嘀嘀无限科技发展有限公司 电动车能耗预测方法、计算机可读存储介质和电子设备
CN111832881A (zh) * 2020-04-30 2020-10-27 北京嘀嘀无限科技发展有限公司 基于路况信息预测电动车能耗的方法、介质和电子设备
WO2022021062A1 (zh) * 2020-07-28 2022-02-03 华为技术有限公司 剩余里程预测的方法和电池远程服务的系统
CN112560186B (zh) * 2020-10-15 2022-08-05 吉林大学 电动汽车行程能耗预测方法、装置、设备及可存储介质
CN112798011B (zh) * 2021-04-15 2021-07-02 天津所托瑞安汽车科技有限公司 车辆里程的计算方法、装置、设备和存储介质
CN113626118B (zh) * 2021-07-30 2023-07-25 中汽创智科技有限公司 能耗实时显示方法、装置及设备
CN113642248B (zh) * 2021-08-30 2023-11-07 平安国际融资租赁有限公司 定位设备剩余使用时间的评估方法及装置
CN114004546A (zh) * 2021-12-30 2022-02-01 南京领行科技股份有限公司 网约车分配方法、装置、电子设备及存储介质
CN117498325B (zh) * 2023-11-02 2024-07-23 深圳中保动力新能源科技有限公司 一种新能源汽车用电量预测及充电资源调度优化方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103213504A (zh) * 2013-04-27 2013-07-24 北京交通大学 一种电动汽车续驶里程估算方法
US20160137090A1 (en) * 2014-11-14 2016-05-19 Hyundai Motor Company System and method for predicting distance to empty of electric vehicle
CN106448137A (zh) * 2016-11-04 2017-02-22 东南大学 基于电动汽车的公交服务系统及方法
CN109726838A (zh) * 2018-12-28 2019-05-07 永安行科技股份有限公司 订单执行方法、执行系统及计算机可读存储介质
CN110222906A (zh) * 2019-06-17 2019-09-10 北京嘀嘀无限科技发展有限公司 电动车能耗预测方法、计算机可读存储介质和电子设备

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3483801A1 (en) * 2017-11-10 2019-05-15 Bayerische Motoren Werke Aktiengesellschaft Methods and apparatuses for fuel consumption prediction
CN109558988B (zh) * 2018-12-13 2021-08-06 北京理工新源信息科技有限公司 一种基于大数据融合的电动汽车能耗预测方法及系统
CN109658203A (zh) * 2018-12-28 2019-04-19 永安行科技股份有限公司 订单分配方法、分配系统及计算机可读存储介质
CN109733248B (zh) * 2019-01-09 2020-07-24 吉林大学 基于路径信息的纯电动汽车剩余里程模型预测方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103213504A (zh) * 2013-04-27 2013-07-24 北京交通大学 一种电动汽车续驶里程估算方法
US20160137090A1 (en) * 2014-11-14 2016-05-19 Hyundai Motor Company System and method for predicting distance to empty of electric vehicle
CN106448137A (zh) * 2016-11-04 2017-02-22 东南大学 基于电动汽车的公交服务系统及方法
CN109726838A (zh) * 2018-12-28 2019-05-07 永安行科技股份有限公司 订单执行方法、执行系统及计算机可读存储介质
CN110222906A (zh) * 2019-06-17 2019-09-10 北京嘀嘀无限科技发展有限公司 电动车能耗预测方法、计算机可读存储介质和电子设备

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115808922A (zh) * 2022-01-07 2023-03-17 宁德时代新能源科技股份有限公司 商用电动车辆能耗预测方法、装置和计算机设备
CN115808922B (zh) * 2022-01-07 2023-10-27 宁德时代新能源科技股份有限公司 商用电动车辆能耗预测方法、装置和计算机设备
CN115061462A (zh) * 2022-05-18 2022-09-16 安徽亚珠金刚石股份有限公司 一种人造金刚石车间用agv车体自动充电规划系统

Also Published As

Publication number Publication date
CN110222906A (zh) 2019-09-10

Similar Documents

Publication Publication Date Title
WO2020253204A1 (zh) 电动车能耗预测方法、计算机可读存储介质和电子设备
CN106908075B (zh) 大数据采集与处理系统及基于其电动汽车续航估计方法
US11703340B2 (en) Trip planning with energy constraint
CN105354896B (zh) 一种汽车行驶记录仪的轨迹分段方法
CN111670340B (zh) 一种车辆剩余行驶里程的获取方法、电子设备及车辆
CN105383496B (zh) 用于车辆的基于路线的剩余能量可行驶距离计算
US9308827B2 (en) Reachable range calculation apparatus, method, and program
US20220383664A1 (en) Systems and methods for providing predictive distance-to-empty for vehicles
CN110174117A (zh) 一种电动汽车充电路线规划方法
CN109784560A (zh) 一种电动汽车续航里程估算方法及估算系统
CN111832881A (zh) 基于路况信息预测电动车能耗的方法、介质和电子设备
WO2014016825A1 (en) Reducing fuel consumption by accommodating to anticipated road and driving conditions
WO2022021062A1 (zh) 剩余里程预测的方法和电池远程服务的系统
CA2840479A1 (en) System and method for generating vehicle drive cycle profiles
US20170010125A1 (en) Device for providing electric-moving-body information and method for providing electric-moving-body information
EP3450919A1 (en) Route estimation apparatus and route estimation method
CN109141459A (zh) 一种带有电耗分析预测的电动汽车导航系统及方法
CN206938676U (zh) 一种用于计算电动汽车续驶里程的装置
CN115817183A (zh) 一种纯电动汽车续驶里程预测方法及预测装置
CN116629425A (zh) 车辆能耗的计算方法、装置、计算机可读介质及电子设备
CN113525385A (zh) 一种车辆行程能耗的预测方法及装置
KR20150008517A (ko) 전기차의 전력 판매 시스템 및 방법
CN115583153A (zh) 一种续航里程计算方法、装置及计算机设备
Fanti et al. An Innovative Service for Electric Vehicle Energy Demand Prediction
CN112406874B (zh) 一种电动汽车远距离充电辅助决策方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19933657

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19933657

Country of ref document: EP

Kind code of ref document: A1