Driving range prediction model construction method, device and system and storage medium

Abstract

The invention discloses a driving range prediction model construction method, device and system and a storage medium, and the method comprises the steps: obtaining first preset factors and energy consumption data at a plurality of sampling moments, the first preset factors comprising at least one preset driving parameter or at least one preset road parameter; determining an association relationship between each first preset factor and the energy consumption data according to the first preset factors at the plurality of sampling moments and the energy consumption data; a plurality of driving range prediction models are constructed according to the incidence relation between the first preset factors and the energy consumption data, and the numbers of the preset factors introduced by the plurality of driving range prediction models are different; and selecting the driving range prediction model of which the prediction error and the calculation speed meet the target requirements from the plurality of driving range prediction models as a finally obtained target driving range prediction model. By adopting the scheme, the accuracy of driving range prediction can be improved.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a method, device, system and storage medium for constructing a driving range prediction model. Background Art

[0002] Range anxiety is a problem that has yet to be fully resolved in the development of pure electric vehicles. The key cause of range anxiety is inaccurate range predictions. Current range prediction methods estimate range based on the remaining battery charge. However, due to the often-variable road conditions on which vehicles travel, and the resulting varying battery consumption, range predictions are often inaccurate. Summary of the Invention

[0003] The present application provides a method, device, system and storage medium for constructing a range prediction model to improve the accuracy of range prediction.

[0004] This application provides a method for constructing a driving range prediction model, comprising:

[0005] Acquiring first preset factors and energy consumption data at multiple sampling moments, where the first preset factors include at least one preset driving parameter or at least one preset road parameter;

[0006] Determining, based on the first preset factors and the energy consumption data at multiple sampling moments, a correlation relationship between each first preset factor and the energy consumption data;

[0007] Constructing a plurality of driving range prediction models according to the correlation between each first preset factor and the energy consumption data, wherein the plurality of driving range prediction models introduce different numbers of preset factors;

[0008] A driving range prediction model whose prediction error and calculation speed both meet target requirements is selected from the multiple driving range prediction models as the final target driving range prediction model.

[0009] The beneficial effects of the present application are as follows: obtaining first preset factors and energy consumption data at multiple sampling moments, wherein the first preset factors include at least one preset driving parameter or at least one preset road parameter; determining the correlation between each first preset factor and the energy consumption data based on the first preset factors and the energy consumption data at multiple sampling moments; constructing multiple driving range prediction models based on the correlation between each first preset factor and the energy consumption data, wherein the multiple driving range prediction models introduce different numbers of preset factors; and selecting a driving range prediction model from the multiple driving range prediction models that meets target requirements in both prediction error and calculation speed as the final target driving range prediction model. Since the preset driving parameters and preset road parameters, two factors that have a significant impact on energy consumption, are introduced when constructing the driving range prediction model, the final driving range prediction model can take into account the driving parameters and road parameters when predicting the driving range, and can maintain a high degree of accuracy even if the route is changed.

[0010] In one embodiment, the method further comprises:

[0011] During the same trip, preset driving parameters, preset road parameters and energy consumption data are obtained according to a preset period;

[0012] storing preset driving parameters, preset road parameters and energy consumption data in a preset list;

[0013] The preset driving parameters, preset road parameters and energy consumption data at the same sampling moment are placed in the same row as the first preset factors and energy consumption data at the same sampling moment;

[0014] The obtaining of the first preset factor and energy consumption data at multiple sampling moments includes:

[0015] The first preset factor and energy consumption data at multiple sampling moments are obtained from the preset list.

[0016] In one embodiment, determining the association relationship between each first preset factor and the energy consumption data based on the first preset factors and the energy consumption data at multiple sampling moments includes:

[0017] Cleaning the first preset factor and energy consumption data to retain the first preset factor and energy consumption data corresponding to trips within a preset distance range;

[0018] The first preset factors after data cleaning are fitted with the energy consumption data to determine the correlation between each first preset factor and the energy consumption data.

[0019] In one embodiment, the constructing of multiple driving range prediction models based on the correlation between each first preset factor and the energy consumption data includes:

[0020] Selecting multiple groups of preset factors from the first preset factors, wherein, during the selection process, the first preset factors having the greatest impact on the energy consumption data are preferentially selected, and the number of preset factors in each group is different;

[0021] According to the correlation between each set of preset factors and the energy consumption data, a driving range prediction model corresponding to each set of preset factors is constructed respectively to obtain multiple driving range prediction models.

[0022] The beneficial effect of this embodiment is that: in the process of selecting multiple groups of preset factors from the first preset factors, the first preset factor with the greatest impact on the energy consumption data is preferentially selected rather than randomly selected, thereby ensuring that the multiple preset factors contained in each selected group of preset factors have the greatest impact on the energy consumption data, thereby further improving the accuracy of the mileage prediction.

[0023] In one embodiment, selecting a target driving range prediction model from the plurality of driving range prediction models whose prediction error and calculation speed both meet target requirements includes:

[0024] Get the mean squared error for each model;

[0025] Obtaining a driving range prediction model with a mean square error less than a preset error;

[0026] The model with the fastest calculation speed is selected from the driving range prediction models whose mean square error is less than the preset error.

[0027] In one embodiment, the method further comprises:

[0028] Receive the destination location transmitted by the vehicle;

[0029] Determine route information between the current location and the destination location;

[0030] Determining a second preset factor corresponding to the route information, where the factor type of the second preset factor is the same as the factor type included in the target driving range prediction model;

[0031] Importing a second preset factor corresponding to the route information into a target driving range prediction model;

[0032] Obtaining a predicted driving range value output by a target driving range prediction model;

[0033] The predicted driving range value is sent to the vehicle end.

[0034] In one embodiment, the method further comprises:

[0035] Obtain actual mileage data sent back by the vehicle;

[0036] Determine the prediction error of the target driving range prediction model based on the actual mileage data and driving range prediction value sent back by the vehicle;

[0037] The machine learning method is used to continue to iterate the model parameters to reduce the prediction error of the target driving range prediction model.

[0038] The present application also provides a device for constructing a driving range prediction model, comprising:

[0039] a first acquisition module, configured to acquire first preset factors and energy consumption data at multiple sampling moments, wherein the first preset factors include at least one preset driving parameter or at least one preset road parameter;

[0040] A first determining module, configured to determine, based on the first preset factors and the energy consumption data at a plurality of sampling moments, an association relationship between each first preset factor and the energy consumption data;

[0041] a construction module, configured to construct a plurality of driving range prediction models according to the correlation between each first preset factor and the energy consumption data, wherein the plurality of driving range prediction models introduce different numbers of preset factors;

[0042] The selection module is used to select a driving range prediction model whose prediction error and calculation speed both meet the target requirements from the multiple driving range prediction models as the final target driving range prediction model.

[0043] In one embodiment, the apparatus further comprises:

[0044] A second acquisition module is used to acquire preset driving parameters, preset road parameters and energy consumption data according to a preset period during the same trip;

[0045] A storage module, used for storing preset driving parameters, preset road parameters and energy consumption data into a preset list;

[0046] The storage module is further used to place the preset driving parameters, preset road parameters and energy consumption data at the same sampling moment in the same row as the first preset factor and energy consumption data at the same sampling moment;

[0047] The first acquisition module is configured to:

[0048] The first preset factor and energy consumption data at multiple sampling moments are obtained from the preset list.

[0049] In one embodiment, the determining module includes:

[0050] a cleaning submodule, configured to clean the first preset factor and the energy consumption data to retain the first preset factor and the energy consumption data corresponding to trips within a preset distance range;

[0051] The fitting submodule is used to fit the first preset factors after data cleaning with the energy consumption data to determine the correlation between each first preset factor and the energy consumption data.

[0052] In one embodiment, the building block comprises:

[0053] A first selection submodule is configured to select multiple groups of preset factors from the first preset factors, wherein, during the selection process, the first preset factors having the greatest impact on the energy consumption data are preferentially selected, and the number of preset factors in each group is different;

[0054] The construction submodule is used to construct a driving range prediction model corresponding to each set of preset factors according to the correlation between each set of preset factors and energy consumption data, so as to obtain multiple driving range prediction models.

[0055] In one embodiment, the selection module includes:

[0056] The first acquisition submodule is used to obtain the mean square error of each model;

[0057] The second acquisition submodule is used to obtain a driving range prediction model with a mean square error less than a preset error;

[0058] The second selection submodule is used to select the model with the fastest calculation speed from the driving range prediction models whose mean square error is less than the preset error.

[0059] In one embodiment, the apparatus further comprises:

[0060] A receiving module is used to receive the destination location transmitted by the vehicle;

[0061] A second determining module is used to determine route information between the current location and the destination location;

[0062] a third determining module, configured to determine a second preset factor corresponding to the route information, wherein the factor type of the second preset factor is the same as the factor type included in the target driving range prediction model;

[0063] An import module, configured to import a second preset factor corresponding to the route information into a target driving range prediction model;

[0064] A third acquisition module is used to obtain the driving range prediction value output by the target driving range prediction model;

[0065] The sending module is used to send the predicted driving range value to the vehicle end.

[0066] In one embodiment, the apparatus further comprises:

[0067] The fourth acquisition module is used to obtain the actual mileage data sent back by the vehicle;

[0068] a fourth determination module, configured to determine a prediction error of a target driving range prediction model based on actual mileage data and a driving range prediction value transmitted back by the vehicle;

[0069] The iteration module is used to continue iterating the model parameters using a machine learning method to reduce the prediction error of the target driving range prediction model.

[0070] This application also provides a driving range prediction model construction system, including:

[0071] at least one processor; and,

[0072] a memory communicatively connected to the at least one processor; wherein,

[0073] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to implement the method for constructing a driving range prediction model described in any of the above embodiments.

[0074] The present application also provides a computer-readable storage medium. When the instructions in the storage medium are executed by a processor corresponding to the driving range prediction model construction system, the driving range prediction model construction system can implement the driving range prediction model construction method described in any of the above embodiments.

[0075] Other features and advantages of the present application will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present application. The purposes and other advantages of the present application can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0076] The technical solution of the present application is further described in detail below through the accompanying drawings and examples. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] The accompanying drawings are used to provide a further understanding of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings:

[0078] Figure 1 This is a flowchart of a method for constructing a driving range prediction model in one embodiment of the present application;

[0079] Figure 2 This is a schematic structural diagram of a device for constructing a driving range prediction model in one embodiment of the present application;

[0080] Figure 3This is a hardware structure diagram of a driving range prediction model construction system in one embodiment of the present application. DETAILED DESCRIPTION

[0081] The preferred embodiments of the present application are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application and are not used to limit the present application.

[0082] Figure 1 This is a flow chart of a method for constructing a driving range prediction model in one embodiment of the present application. Figure 1 As shown, the method can be implemented as the following steps S101-S104:

[0083] In step S101, a first preset factor and energy consumption data at a plurality of sampling moments are obtained, wherein the first preset factor includes at least one preset driving parameter or at least one preset road parameter;

[0084] In step S102, the correlation relationship between each first preset factor and the energy consumption data is determined based on the first preset factors and the energy consumption data at multiple sampling moments;

[0085] In step S103, a plurality of driving range prediction models are constructed based on the correlation between each first preset factor and the energy consumption data, wherein the plurality of driving range prediction models introduce different numbers of preset factors;

[0086] In step S104, a driving range prediction model whose prediction error and calculation speed both meet target requirements is selected from the multiple driving range prediction models as the final target driving range prediction model.

[0087] Acquire a first preset factor and energy consumption data at multiple sampling moments. The first preset factor includes at least one preset driving parameter or at least one preset road parameter. In this application, to improve the accuracy of range prediction, preset driving parameters, preset road parameters, and energy consumption data are acquired during a trip according to a preset period. Preset driving parameters include mileage S and vehicle speed v. Preset road parameters include the number of traffic lights, average road slope, road grade, and traffic flow conditions. Road grades include expressways, national highways, provincial highways, county roads, and urban expressways, and traffic flow conditions include extremely smooth, smooth, normal, congested, and extremely congested. Road grades and traffic flow conditions are pre-established and stored. Preset road parameters may also include environmental information such as temperature, humidity, altitude, weather, wind speed, and wind direction. Vehicle energy consumption data includes battery level E and state of charge (SOC). During data acquisition, the above data is collected according to a preset period, for example, a 1s acquisition period. In addition, the preset driving parameters, preset road parameters and energy consumption data are stored in a preset list; during the storage process, the preset driving parameters, preset road parameters and energy consumption data at the same sampling moment are placed in the same row as the first preset factors and energy consumption data at the same sampling moment, and then, the first preset factors and energy consumption data at multiple sampling moments can be obtained from the preset list based on the sampling moment.

[0088] For example, for a trip from point A to point B, which takes 120 seconds and covers 2 km, if the collection period is 1 second, the driving data will include 120 time points, recording the preset driving parameters corresponding to each time point: mileage S, speed v, vehicle energy consumption data (electricity E, SOC), and preset road parameters, such as an average road slope of 5%, a national highway grade, smooth traffic flow, a driving temperature of 20°C, a humidity of 50%, an altitude of 1000 meters, sunny weather, a wind speed of level 3, and a southeast wind direction.

[0089] The correlation between each first preset factor and the energy consumption data is determined based on the first preset factors and the energy consumption data at multiple sampling moments. (1) Data cleaning: In order to ensure the accuracy of the data, a preset driving range is set, for example, 1km-500km. The current driving distance is determined by the difference between the driving mileage corresponding to different sampling moments and the driving mileage at the initial moment. When the driving distance is within the preset driving range, the data at the current moment is determined to be valid data, and then the first preset factor and energy consumption data are cleaned to retain the first preset factor and energy consumption data corresponding to the trip within the preset distance range. It is understandable that different preset driving ranges can also be set according to different road grades, and samples that exceed the preset driving range are eliminated, thereby achieving cleaning of samples of different road grades. For example, when determining the driving distance on a county road, the difference between the driving mileage corresponding to the time of entering the county road and the driving mileage when leaving the county road is used as the distance of the county road. Assuming that the preset driving range of the county road is 1-20km, the data of driving 100km can be eliminated. (2) Data fitting: After data cleaning, the first preset factors after data cleaning are fitted with the energy consumption data to determine the correlation between each first preset factor and the energy consumption data. Before data fitting, in order to obtain sample data for building the model, the data can also be preprocessed. Specifically, ① first assign values ​​to non-numerical data, such as road grade, traffic flow, weather, wind speed, wind direction, etc. according to the preset assignment rules, for example, the road grades of expressway, national highway, provincial highway, county highway, urban expressway, and urban ordinary road are assigned values ​​of 10, 9, 8, 7, 6, and 5 respectively; ② Then, the collected data is dimensionless. During dimensionless transformation, since the non-numerical data has been dimensionless when assigned, no processing is required. For numerical data, data standardization is performed based on the average value of the collected data to obtain the preprocessed first preset factors and the processed energy consumption data. ③ In addition, the first preset factor data and energy consumption data can be further processed to obtain the first preset factor change value and energy consumption data change value corresponding to the target road section. Specifically, the driving time corresponding to the start of the trip to the current sampling moment is determined by subtracting the current sampling moment of the vehicle from the initial driving moment of this trip (i.e., the first initial moment); the driving distance, power change and consumption SOC change corresponding to the current sampling moment of the vehicle are determined by the difference between the current mileage, power E, SOC of the vehicle and the mileage, power E, SOC at the initial moment.Furthermore, in order to obtain data under different road conditions, when the vehicle enters the target road section, the corresponding sampling moment is marked as the second initial moment, and the time when the vehicle leaves the preset target road section is subtracted from the second initial moment to obtain the vehicle's travel time on the corresponding road; the difference between the vehicle's mileage, battery charge E, and SOC when leaving the target road section and the mileage, battery charge E, and SOC at the initial moment is used to determine the vehicle's travel distance, battery charge change, and SOC change on the target road section. ④ Finally, the correlation relationship between each first preset factor and the energy consumption data is determined based on the first preset factor of the target road section, the change value of the first preset factor, the energy consumption data, and the change value of the energy consumption data. The correlation relationship includes a correlation degree and a functional relationship. For example, to obtain the relationship between each first preset factor and energy consumption data on a highway, each first preset factor and / or first preset factor change value and energy consumption data and / or energy consumption data change value at multiple sampling moments corresponding to the highway can be extracted as sample data, and the data correlation between the processed first preset factors and the energy consumption data can be determined through gray correlation analysis, and the degree of correlation between the data can be measured by the magnitude of the correlation. For example, energy consumption change data can be obtained, including power change ΔE and SOC change ΔSOC; then, the correlation between each first preset factor and ΔE and ΔSOC can be calculated. The degree of influence of each first preset factor on the change in energy consumption data can be determined by the magnitude of the correlation, with the greater the correlation, the greater the degree of influence. Of course, the functional relationship between the first preset factors and energy consumption data can also be determined through methods such as correlation analysis. For example, the correlation ranking is vehicle speed > temperature > number of traffic lights, that is, the vehicle speed v has the greatest correlation with the power change ΔE, and further fitting results in ΔE≈3*v. 2.7 .

[0090] In addition, when determining the correlation between each first preset factor and the energy consumption data based on the first preset factors and the energy consumption data at multiple sampling moments, the first preset factor can also be input into a specific model, and then the model will output the energy consumption result based on the first preset factor, and then establish the correlation between the first preset factor and the energy consumption result.

[0091] Multiple driving range prediction models are constructed based on the correlations between each first preset factor and the energy consumption data, wherein the multiple driving range prediction models incorporate different numbers of preset factors. Specifically, multiple groups of preset factors are selected from the first preset factors. During the selection process, the first preset factors with the greatest impact on the energy consumption data are prioritized, and each group of preset factors has a different number of preset factors. Since a prediction model with more factors increases the computational complexity, the computational error gradually decreases with the increase in the number of factors and then stabilizes. Therefore, factors are ranked by correlation, and selected in descending order of correlation, such as the top 5 factors, the top 10 factors, the top 20 factors, and so on. Then, based on the correlations between each group of preset factors and the energy consumption data, a driving range prediction model corresponding to each group of preset factors is constructed, thereby obtaining multiple driving range prediction models. For example, if the top 5 factors are selected, prediction models corresponding to the five factors are constructed based on the correlations between the first preset factors and the energy consumption data. The multiple groups of data corresponding to the five factors are then input into the prediction model for fitting with the energy consumption data to obtain the prediction model for the corresponding factor. Of course, it is also possible to establish a prediction model for the corresponding factors through machine learning algorithms, such as neural network models.

[0092] A range prediction model whose prediction error and computational speed both meet the target requirements is selected from the multiple range prediction models as the final target range prediction model. The mean square error (MSE) of each model is obtained. The numerical values ​​of multiple test sample groups are input into the prediction model to obtain corresponding predicted values. Based on the predicted values ​​and the corresponding measured values, the MSE and computational runtime of each model are obtained. Next, a range prediction model whose MSE is less than a preset error is obtained. Among the range prediction models whose MSE is less than the preset error, the model with the fastest computational speed is selected as the final selected range prediction model. The final range prediction model ensures that the prediction error requirement is met (reducing the range prediction error) while maintaining a sufficiently fast computational speed (reducing the computational resources required to run the model).

[0093] Furthermore, when applying the model, the ΔE calculated by the model can be converted into ΔSOC. By subtracting the consumed ΔSOC from the vehicle's current real-time SOC, the remaining available SOC upon arrival at the destination can be calculated. For example, if the SOC at departure is 70%, the model, based on a preset route and the distance traveled on different roads, predicts that the trip will require 30kWh of power. This is converted to a consumed SOC of 30%, and the remaining available SOC upon arrival at the destination is calculated as 70% - 30% = 40%. This means that the user will have 40% SOC remaining upon arrival, indicating that the battery is sufficient to support the user's journey to the destination.

[0094] The model established in this application can be deployed in the cloud, which will then send the corresponding model to the onboard computing unit (i.e., the vehicle side) of the corresponding vehicle model. The user will send the corresponding form data back to the cloud at the beginning and end of each formation. The cloud will iterate the model based on the data of the corresponding vehicle model according to a preset period, or when the model calculation error and the actual error are greater than a preset value. The iterated model can be resent to the vehicle side for operation after verification.

[0095] Of course, to save cloud computing power, the target range prediction model can also be deployed on-board. That is, the model utilizes edge computing principles, with each vehicle equipped with the model performing on-board computations. In this case, the solution can also be implemented by transmitting the target range prediction model to the vehicle. When the target range prediction model operates on-board, it operates as follows: upon receiving the destination location transmitted by the navigation program, the model determines the route between the current location and the destination; determines a second preset factor corresponding to the route information, the second preset factor being of the same type as the factor included in the target range prediction model; imports the second preset factor corresponding to the route information into the target range prediction model; and obtains the target range prediction value output by the target range prediction model. Because the model utilizes edge computing principles, with each vehicle equipped with the model performing on-board computations, it reduces network bandwidth consumption and cloud computing power consumption, and ensures that computations are not affected by temporary network signal loss. When a user begins a trip, they first need to set their destination in the navigation program and select a route. Relevant information is input into the model as calculation factors. Vehicle-related information, such as battery level E and SOC, is also factored into the model. The model performs calculations at a frequency of Y, for example, Y = once every 10 seconds. This frequency is customizable and determines how often the prediction results are updated during the journey. For example, if a traffic accident occurs during the journey and the previously smooth road becomes congested, the model can obtain real-time traffic information and update the prediction results, providing users with the most accurate range information in real time. This is a feature that traditional range models have not been able to achieve.

[0096] The beneficial effects of the present application are as follows: obtaining first preset factors and energy consumption data at multiple sampling moments, wherein the first preset factors include at least one preset driving parameter or at least one preset road parameter; determining the correlation between each first preset factor and the energy consumption data based on the first preset factors and the energy consumption data at multiple sampling moments; constructing multiple driving range prediction models based on the correlation between each first preset factor and the energy consumption data, wherein the multiple driving range prediction models introduce different numbers of preset factors; and selecting a driving range prediction model from the multiple driving range prediction models that meets target requirements in both prediction error and calculation speed as the final target driving range prediction model. Since the preset driving parameters and preset road parameters, two factors that have a significant impact on energy consumption, are introduced when constructing the driving range prediction model, the final driving range prediction model can take into account the driving parameters and road parameters when predicting the driving range, and can maintain a high degree of accuracy even if the route is changed.

[0097] In one embodiment, the above method may also be implemented as the following steps A1-A3:

[0098] In step A1, during the same trip, preset driving parameters, preset road parameters, and energy consumption data are obtained according to a preset period;

[0099] In step A2, the preset driving parameters, preset road parameters and energy consumption data are stored in a preset list;

[0100] In step A3, the preset driving parameters, preset road parameters, and energy consumption data at the same sampling moment are placed in the same row as the first preset factors and energy consumption data at the same sampling moment;

[0101] The above step S101 can be implemented as the following step A4:

[0102] In step A4, the first preset factor and energy consumption data at multiple sampling moments are obtained from the preset list.

[0103] In one embodiment, the above step S102 may be implemented as the following steps B1-B2:

[0104] In step B1, the first preset factor and energy consumption data are cleaned to retain the first preset factor and energy consumption data corresponding to trips within a preset distance range;

[0105] In step B2, the first preset factors after data cleaning are fitted with the energy consumption data to determine the correlation between each first preset factor and the energy consumption data.

[0106] In one embodiment, the above step S103 may be implemented as the following steps C1-C2:

[0107] In step C1, multiple groups of preset factors are selected from the first preset factors, wherein, during the selection process, the first preset factors having the greatest impact on the energy consumption data are preferentially selected, and the number of preset factors in each group is different;

[0108] In step C2, a driving range prediction model corresponding to each set of preset factors is constructed according to the correlation between each set of preset factors and the energy consumption data, so as to obtain a plurality of driving range prediction models.

[0109] The beneficial effect of this embodiment is that: in the process of selecting multiple groups of preset factors from the first preset factors, the first preset factor with the greatest impact on the energy consumption data is preferentially selected rather than randomly selected, thereby ensuring that the multiple preset factors contained in each selected group of preset factors have the greatest impact on the energy consumption data, thereby further improving the accuracy of the mileage prediction.

[0110] In one embodiment, the above step S104 may be implemented as the following steps D1-D3:

[0111] In step D1, the mean square error of each model is obtained;

[0112] In step D2, a driving range prediction model is obtained whose mean square error is less than a preset error;

[0113] In step D3, a model with the fastest calculation speed is selected from the range prediction models whose mean square error is less than a preset error.

[0114] In one embodiment, the above method may also be implemented as the following steps E1-E2:

[0115] In step E1, the destination location transmitted by the vehicle is received;

[0116] In step E2, route information between the current location and the destination location is determined;

[0117] In step E3, a second preset factor corresponding to the route information is determined, where the factor type of the second preset factor is the same as the factor type included in the target driving range prediction model;

[0118] In step E4, a second preset factor corresponding to the route information is imported into a target driving range prediction model;

[0119] In step E5, a driving range prediction value output by the target driving range prediction model is obtained;

[0120] In step E6, the predicted driving range value is sent to the vehicle side.

[0121] In one embodiment, the above method may also be implemented as the following steps F1-F2:

[0122] In step F1, the actual mileage data transmitted by the vehicle is obtained;

[0123] In step F2, the prediction error of the target driving range prediction model is determined based on the actual mileage data and the driving range prediction value transmitted back by the vehicle;

[0124] In step F3, the model parameters are further iterated using a machine learning method to reduce the prediction error of the target driving range prediction model.

[0125] In one embodiment of the present application, the driving range prediction model is deployed on the vehicle side, and the working process of the target driving range prediction model on the vehicle side can be implemented as follows: Steps G1-G4:

[0126] In step G1, upon receiving the destination location transmitted by the navigation program, determining route information between the current location and the destination location;

[0127] In step G2, a second preset factor corresponding to the route information is determined, where the factor type of the second preset factor is the same as the factor type included in the target driving range prediction model;

[0128] In step G3, a second preset factor corresponding to the route information is imported into a target driving range prediction model;

[0129] In step G4, a driving range prediction value output by the target driving range prediction model is obtained.

[0130] In this embodiment, to conserve cloud computing power, the target range prediction model can also be deployed on the vehicle side. That is, the model utilizes edge computing principles, with each vehicle equipped with the model performing on-vehicle computations. In this case, the solution can also be implemented by transmitting the target range prediction model to the vehicle side. In this case, the target range prediction model operates as follows on the vehicle side: upon receiving the destination location transmitted by the navigation program, the model determines the route information between the current location and the destination location; determines a second preset factor corresponding to the route information, where the factor type of the second preset factor is the same as the factor type included in the target range prediction model; imports the second preset factor corresponding to the route information into the target range prediction model; and obtains the range prediction value output by the target range prediction model.

[0131] Figure 2 FIG. 1 is a structural diagram of a device for constructing a driving range prediction model in one embodiment of the present application. Figure 2 As shown, the device includes:

[0132] A first acquisition module 201 is configured to acquire a first preset factor and energy consumption data at a plurality of sampling moments, wherein the first preset factor includes at least one preset driving parameter or at least one preset road parameter;

[0133] A first determining module 202 is configured to determine, based on the first preset factors and the energy consumption data at multiple sampling moments, an association relationship between each first preset factor and the energy consumption data;

[0134] A construction module 203 is configured to construct a plurality of driving range prediction models based on the correlation between each first preset factor and the energy consumption data, wherein the plurality of driving range prediction models include different numbers of preset factors;

[0135] The selection module 204 is configured to select a driving range prediction model whose prediction error and calculation speed both meet target requirements from the plurality of driving range prediction models as a final target driving range prediction model.

[0136] In one embodiment, the apparatus further comprises:

[0137] A second acquisition module is used to acquire preset driving parameters, preset road parameters and energy consumption data according to a preset period during the same trip;

[0138] A storage module, used for storing preset driving parameters, preset road parameters and energy consumption data into a preset list;

[0139] The storage module is further used to place the preset driving parameters, preset road parameters and energy consumption data at the same sampling moment in the same row as the first preset factor and energy consumption data at the same sampling moment;

[0140] The first acquisition module is configured to:

[0141] The first preset factor and energy consumption data at multiple sampling moments are obtained from the preset list.

[0142] In one embodiment, the determining module includes:

[0143] a cleaning submodule, configured to clean the first preset factor and the energy consumption data to retain the first preset factor and the energy consumption data corresponding to trips within a preset distance range;

[0144] The fitting submodule is used to fit the first preset factors after data cleaning with the energy consumption data to determine the correlation between each first preset factor and the energy consumption data.

[0145] In one embodiment, the building block comprises:

[0146] A first selection submodule is configured to select multiple groups of preset factors from the first preset factors, wherein, during the selection process, the first preset factors having the greatest impact on the energy consumption data are preferentially selected, and the number of preset factors in each group is different;

[0147] The construction submodule is used to construct a driving range prediction model corresponding to each set of preset factors according to the correlation between each set of preset factors and energy consumption data, so as to obtain multiple driving range prediction models.

[0148] In one embodiment, the selection module includes:

[0149] The first acquisition submodule is used to obtain the mean square error of each model;

[0150] The second acquisition submodule is used to obtain a driving range prediction model with a mean square error less than a preset error;

[0151] The second selection submodule is used to select the model with the fastest calculation speed from the driving range prediction models whose mean square error is less than the preset error.

[0152] In one embodiment, the apparatus further comprises:

[0153] A receiving module is used to receive the destination location transmitted by the vehicle;

[0154] A second determining module is used to determine route information between the current location and the destination location;

[0155] a third determining module, configured to determine a second preset factor corresponding to the route information, wherein the factor type of the second preset factor is the same as the factor type included in the target driving range prediction model;

[0156] An import module, configured to import a second preset factor corresponding to the route information into a target driving range prediction model;

[0157] A third acquisition module is used to obtain the driving range prediction value output by the target driving range prediction model;

[0158] The sending module is used to send the predicted driving range value to the vehicle end.

[0159] In one embodiment, the apparatus further comprises:

[0160] The fourth acquisition module is used to obtain the actual mileage data sent back by the vehicle;

[0161] a fourth determination module, configured to determine a prediction error of a target driving range prediction model based on actual mileage data and a driving range prediction value transmitted back by the vehicle;

[0162] The iteration module is used to continue iterating the model parameters using a machine learning method to reduce the prediction error of the target driving range prediction model.

[0163] This application also provides a driving range prediction model construction system, including:

[0164] at least one processor; and,

[0165] a memory communicatively connected to the at least one processor; wherein,

[0166] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to implement the method for constructing a driving range prediction model described in any of the above embodiments.

[0167] Figure 3 FIG. 1 is a schematic diagram of the hardware structure of a driving range prediction model construction system in one embodiment of the present application. Figure 3 As shown, the driving range prediction model construction system includes:

[0168] at least one processor 320; and,

[0169] A memory 304 in communication with the at least one processor 320; wherein,

[0170] The memory 304 stores instructions that can be executed by the at least one processor 320, and the instructions are executed by the at least one processor 320 to implement the method for constructing a driving range prediction model described in any of the above embodiments.

[0171] Reference Figure 3 The driving range prediction model building system 300 may include one or more of the following components: a processing component 302 , a memory 304 , a power component 306 , a multimedia component 308 , an audio component 310 , an input / output (I / O) interface 312 , a sensor component 314 , and a communication component 316 .

[0172] The processing component 302 generally controls the overall operation of the range prediction model building system 300. The processing component 302 may include one or more processors 320 to execute instructions to complete all or part of the steps of the aforementioned method. Furthermore, the processing component 302 may include one or more modules to facilitate interaction between the processing component 302 and other components. For example, the processing component 302 may include a multimedia module to facilitate interaction between the multimedia component 308 and the processing component 302.

[0173] The memory 304 is configured to store various types of data to support the operation of the driving range prediction model building system 300. Examples of such data include instructions for any application or method operating on the driving range prediction model building system 300, such as text, images, videos, etc. The memory 304 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0174] The power supply component 306 provides power to various components of the range prediction model building system 300. The power supply component 306 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the range prediction model building system 300.

[0175] The multimedia component 308 includes a screen that provides an output interface between the driving range prediction model building system 300 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, slides, and gestures on the touch panel. The touch sensor can not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 308 may also include a front camera and / or a rear camera. When the driving range prediction model building system 300 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have focal length and optical zoom capabilities.

[0176] The audio component 310 is configured to output and / or input audio signals. For example, the audio component 310 includes a microphone (MIC) that is configured to receive external audio signals when the range prediction model building system 300 is in an operating mode, such as an alarm mode, a recording mode, a voice recognition mode, and a voice output mode. The received audio signal can be further stored in the memory 304 or transmitted via the communication component 316. In some embodiments, the audio component 310 also includes a speaker for outputting audio signals.

[0177] I / O interface 312 provides an interface between processing component 302 and peripheral interface modules, such as a keyboard, click wheel, buttons, etc. These buttons may include but are not limited to: a home button, volume buttons, a start button, and a lock button.

[0178] Sensor assembly 314 includes one or more sensors for providing various status assessments for range prediction model building system 300. For example, sensor assembly 314 may include an acoustic sensor. Additionally, sensor assembly 314 may detect the on / off status of range prediction model building system 300, the relative positioning of components, such as the display and keypad of range prediction model building system 300, and the operating status of range prediction model building system 300 or a component thereof, such as the operating status, structural status, or operating status of a discharge scraper, the orientation or acceleration / deceleration of range prediction model building system 300, and temperature changes within range prediction model building system 300. Sensor assembly 314 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 314 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 314 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, a material stack thickness sensor, or a temperature sensor.

[0179] The communication component 316 is configured to enable the driving range prediction model building system 300 to provide the ability to communicate with other devices and cloud platforms in a wired or wireless manner. The driving range prediction model building system 300 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 316 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 316 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0180] In an exemplary embodiment, the driving range prediction model construction system 300 can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to execute the driving range prediction model construction method described in any of the above embodiments.

[0181] The present application also provides a computer-readable storage medium. When the instructions in the storage medium are executed by a processor corresponding to the driving range prediction model construction system, the driving range prediction model construction system can implement the driving range prediction model construction method described in any of the above embodiments.

[0182] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage and optical storage, etc.) that contain computer-usable program code.

[0183] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0184] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0185] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0186] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A method for constructing a driving range prediction model, characterized in that: include: Acquiring first preset factors and energy consumption data at multiple sampling moments, where the first preset factors include at least one preset driving parameter or at least one preset road parameter; Determining, based on the first preset factors and the energy consumption data at multiple sampling moments, a correlation relationship between each first preset factor and the energy consumption data; Constructing a plurality of driving range prediction models according to the correlation between each first preset factor and the energy consumption data, wherein the plurality of driving range prediction models introduce different numbers of preset factors; A driving range prediction model whose prediction error and calculation speed both meet target requirements is selected from the multiple driving range prediction models as the final target driving range prediction model.

2. The method according to claim 1, wherein The method further comprises: During the same trip, preset driving parameters, preset road parameters and energy consumption data are obtained according to a preset period; storing preset driving parameters, preset road parameters and energy consumption data in a preset list; The preset driving parameters, preset road parameters and energy consumption data at the same sampling moment are placed in the same row as the first preset factors and energy consumption data at the same sampling moment; The obtaining of the first preset factor and energy consumption data at multiple sampling moments includes: The first preset factor and energy consumption data at multiple sampling moments are obtained from the preset list.

3. The method according to claim 1, wherein The determining of the correlation between each first preset factor and the energy consumption data according to the first preset factors and the energy consumption data at multiple sampling moments includes: Cleaning the first preset factor and energy consumption data to retain the first preset factor and energy consumption data corresponding to trips within a preset distance range; The first preset factors after data cleaning are fitted with the energy consumption data to determine the correlation between each first preset factor and the energy consumption data.

4. The method according to claim 1, wherein The method of constructing multiple driving range prediction models based on the correlation between each first preset factor and the energy consumption data includes: Selecting multiple groups of preset factors from the first preset factors, wherein, during the selection process, the first preset factors having the greatest impact on the energy consumption data are preferentially selected, and the number of preset factors in each group is different; According to the correlation between each set of preset factors and the energy consumption data, a driving range prediction model corresponding to each set of preset factors is constructed respectively to obtain multiple driving range prediction models.

5. The method according to claim 1, wherein The step of selecting a target driving range prediction model from the plurality of driving range prediction models, wherein both the prediction error and the calculation speed meet target requirements, comprises: Get the mean squared error for each model; Obtaining a driving range prediction model with a mean square error less than a preset error; The model with the fastest calculation speed is selected from the driving range prediction models whose mean square error is less than the preset error.

6. The method according to claim 1, wherein The method further comprises: Receive the destination location transmitted by the vehicle; Determine route information between the current location and the destination location; Determining a second preset factor corresponding to the route information, where the factor type of the second preset factor is the same as the factor type included in the target driving range prediction model; Importing a second preset factor corresponding to the route information into a target driving range prediction model; Obtaining a predicted driving range value output by a target driving range prediction model; The predicted driving range value is sent to the vehicle end.

7. The method according to claim 1, wherein The method further comprises: Obtain actual mileage data sent back by the vehicle; Determine the prediction error of the target driving range prediction model based on the actual mileage data and driving range prediction value sent back by the vehicle; The machine learning method is used to continue to iterate the model parameters to reduce the prediction error of the target driving range prediction model.

8. A device for constructing a driving range prediction model, characterized in that: include: a first acquisition module, configured to acquire first preset factors and energy consumption data at multiple sampling moments, wherein the first preset factors include at least one preset driving parameter or at least one preset road parameter; A first determining module, configured to determine, based on the first preset factors and the energy consumption data at a plurality of sampling moments, an association relationship between each first preset factor and the energy consumption data; a construction module, configured to construct a plurality of driving range prediction models according to the correlation between each first preset factor and the energy consumption data, wherein the plurality of driving range prediction models introduce different numbers of preset factors; The selection module is used to select a driving range prediction model whose prediction error and calculation speed both meet the target requirements from the multiple driving range prediction models as the final target driving range prediction model.

9. A driving range prediction model construction system, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to implement the method for constructing a driving range prediction model as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that When the instructions in the storage medium are executed by a processor corresponding to the driving range prediction model construction system, the driving range prediction model construction system can implement the driving range prediction model construction method according to any one of claims 1 to 7.

Images