Range prediction method, device, equipment, medium and vehicle

Abstract

The application discloses a kind of endurance mileage prediction method, device, equipment, medium and vehicle, belong to machine learning technical field.The method comprises: the battery state of charge SOC of target vehicle is obtained;In the SOC of target vehicle greater than or equal to the first threshold value, the first vehicle data of target vehicle is input to the first endurance mileage prediction model, the endurance mileage of target vehicle is predicted using the first endurance mileage prediction model, and the first endurance mileage of pure electricity is obtained;In the SOC of target vehicle less than the first threshold value, the second vehicle data of target vehicle is input to the second endurance mileage prediction model, the endurance mileage of target vehicle is predicted using the second endurance mileage prediction model, and the second endurance mileage of pure electricity is obtained.The endurance mileage prediction method, device, equipment, medium and vehicle provided in the application can improve the accuracy of endurance mileage prediction of pure electricity.

Description

Technical Field

[0001] This application belongs to the field of deep learning technology, and in particular relates to a method, device, equipment, medium and vehicle for predicting driving range. Background Technology

[0002] With the increasing popularity of new energy vehicles, the requirements for the accuracy of pure electric range prediction of new energy vehicles are becoming higher and higher in order to improve users' driving experience.

[0003] In existing technologies, the pure electric driving range of a vehicle is usually predicted based on the New European Driving Cycle (NEDC) / World Light Vehicle Test Procedure (WLTC) conditions.

[0004] However, since the NEDC / WLTC cycle represents ideal conditions, the predicted pure electric range of a vehicle based on the NEDC / WLTC cycle will decrease regularly with a fixed coefficient. In reality, the pure electric range of a vehicle is affected by many factors and will not decrease regularly with a fixed coefficient. Therefore, the prediction of the pure electric range of a vehicle based on the NEDC / WLTC cycle is not accurate. Summary of the Invention

[0005] This application provides a method, apparatus, device, medium, and vehicle for predicting driving range, which can at least solve the problem of inaccurate prediction of pure electric driving range in the prior art.

[0006] In a first aspect, embodiments of this application provide a method for improving pure electric driving range, the method comprising:

[0007] Obtain the battery state of charge (SOC) of the target vehicle;

[0008] When the SOC of the target vehicle is greater than or equal to the first threshold, the first vehicle data of the target vehicle is input into the first range prediction model. The first range prediction model is used to predict the pure electric range of the target vehicle to obtain the first pure electric range. The first vehicle data includes vehicle driving data, vehicle environmental data and battery status data. The first range prediction model is trained based on the first historical vehicle data of the first vehicle with the same model as the target vehicle. The first historical vehicle data is obtained when the SOC of the first vehicle is greater than or equal to the first threshold.

[0009] When the SOC of the target vehicle is less than the first threshold, the second vehicle data of the target vehicle is input into the second range prediction model. The second range prediction model is used to predict the pure electric range of the target vehicle to obtain the second pure electric range. The second vehicle data includes the SOC of the target vehicle, the maximum temperature of the battery pack, and the minimum temperature of the battery pack. The second range prediction model is trained based on the second historical vehicle data of the second vehicle with the same model as the target vehicle. The second historical vehicle data is obtained when the SOC of the second vehicle is less than the first threshold.

[0010] Secondly, embodiments of this application provide a driving range prediction device, the device comprising:

[0011] The first acquisition module is used to acquire the battery state of charge (SOC) of the target vehicle.

[0012] The first prediction module is used to input the first vehicle data of the target vehicle into the first range prediction model when the SOC is greater than or equal to the first threshold, and use the first range prediction model to predict the pure electric range of the target vehicle to obtain the first pure electric range. The first vehicle data includes vehicle driving data, vehicle environment data and battery status data. The first range prediction model is trained based on the first historical vehicle data of the first vehicle with the same model as the target vehicle. The first historical vehicle data is obtained when the SOC of the first vehicle is greater than or equal to the first threshold.

[0013] The second prediction module is used to input the second vehicle data of the target vehicle into the second range prediction model when the SOC is less than the first threshold, and use the second range prediction model to predict the pure electric range of the target vehicle to obtain the second pure electric range. The second vehicle data includes the SOC of the target vehicle, the maximum temperature of the battery pack, and the minimum temperature of the battery pack. The second range prediction model is trained based on the second historical vehicle data of the second vehicle with the same model as the target vehicle. The second historical vehicle data is obtained when the SOC of the second vehicle is less than the first threshold.

[0014] Thirdly, embodiments of this application provide an electronic device, the device comprising: a processor and a memory storing computer program instructions;

[0015] When the processor executes the computer program instructions, it implements the range prediction method as shown in any embodiment of the first aspect.

[0016] Fourthly, embodiments of this application provide a computer storage medium storing computer program instructions, which, when executed by a processor, implement the range prediction method shown in any embodiment of the first aspect.

[0017] Fifthly, embodiments of this application also provide a vehicle, the vehicle including: a range prediction device as shown in any embodiment of the second aspect or an electronic device as shown in any embodiment of the third aspect.

[0018] The driving range prediction method, apparatus, device, medium, and vehicle of this application embodiment can predict the pure electric driving range of a target vehicle using a first driving range prediction model when the target vehicle's State of Charge (SOC) is greater than or equal to a first threshold, and predict the pure electric driving range of the target vehicle using a second driving range prediction model when the target vehicle's SOC is less than the first threshold. This allows for prediction of the vehicle's pure electric driving range using a machine learning model, resulting in a more accurate prediction. Furthermore, the first driving range prediction model is trained based on first historical vehicle data of a first vehicle of the same model as the target vehicle, obtained when the first vehicle's SOC is greater than or equal to the first threshold. The second driving range prediction model is trained based on second historical vehicle data of a second vehicle of the same model as the target vehicle, obtained when the second vehicle's SOC is less than the first threshold. Using different driving range prediction models for vehicles with different SOCs further improves the accuracy of pure electric driving range prediction.

[0019] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of a driving range prediction method provided in one embodiment of this application;

[0022] Figure 2 This is a flowchart of another driving range prediction method provided in one embodiment of this application;

[0023] Figure 3 This is a schematic diagram of the structure of a driving range prediction device provided in one embodiment of this application;

[0024] Figure 4 This is a schematic diagram of the structure of an electronic device provided in one embodiment of this application. Detailed Implementation

[0025] To better understand the above-mentioned objectives, features, and advantages of this application, the solution of this application will be further described below. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0026] Many specific details are set forth in the following description in order to provide a full understanding of this application, but this application may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some embodiments of this application, and not all embodiments.

[0027] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the term "comprising" or any other variations thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0028] As mentioned in the background section, existing technologies typically predict a vehicle's pure electric range based on the NEDC / WLTC driving cycle. However, since the NEDC / WLTC cycle represents ideal conditions, the predicted pure electric range will decrease systematically with a fixed coefficient. In reality, however, a vehicle's pure electric range is affected by various factors and does not decrease systematically with a fixed coefficient. Therefore, predicting a vehicle's pure electric range based on the NEDC / WLTC cycle is inaccurate.

[0029] Other existing methods for predicting pure electric range may result in an overall overestimation of the predicted pure electric range when fully charged. As the vehicle drives and the battery discharges, the range decreases at a rate close to the actual value in the high state of charge (SOC) range. Therefore, the predicted pure electric range in the high SOC range is also generally overestimated. However, when in the low SOC range, the predicted pure electric range decreases rapidly, causing range anxiety for users.

[0030] Based on this, embodiments of this application provide a driving range prediction method. When the target vehicle's State of Charge (SOC) is greater than or equal to a first threshold, a first driving range prediction model is used to predict the pure electric driving range of the target vehicle. When the target vehicle's SOC is less than the first threshold, a second driving range prediction model is used to predict the pure electric driving range of the target vehicle. In this way, a machine learning model can be used to predict the vehicle's pure electric driving range, making the predicted pure electric driving range more accurate. Furthermore, using different driving range prediction models for vehicles with different SOCs can further improve the accuracy of the pure electric driving range prediction.

[0031] Figure 1 The diagram illustrates a flowchart of a range prediction method according to an embodiment of this application. Specifically, it may be a training method for the first range prediction model and the second range prediction model used in the range prediction method provided in an embodiment of this application.

[0032] like Figure 1 As shown, the training methods for the first and second range prediction models used in the range prediction method provided in this application embodiment may include S110-S120:

[0033] S110, acquire multiple first training samples and multiple second training samples;

[0034] S120: Train a first range prediction model based on multiple first training samples until the first range prediction model converges to obtain a trained first range prediction model; and train a second range prediction model based on multiple second training samples until the second range prediction model converges to obtain a trained second range prediction model.

[0035] Therefore, by training a first range prediction model using a first training sample constructed from the first historical data of a first vehicle with the same model as the target vehicle, and training a second range prediction model using a second training sample constructed from the second historical data of a second vehicle with the same model as the target vehicle, a well-trained first and second range prediction model can be obtained. Thus, the pure electric range of the target vehicle can be predicted more accurately using the well-trained first and second range prediction models.

[0036] Regarding S110, the data required to predict the pure electric range may differ for different vehicle models. Therefore, the data in the first and second training samples used to train the first and second range prediction models can be the history of vehicles of the same model as the target vehicle, where the target vehicle can be the vehicle whose pure electric range is to be predicted.

[0037] The first training sample may include first historical data of a first vehicle with the same model as the target vehicle, and the second training sample may include second historical data of a second vehicle with the same model as the target vehicle. The first historical data may include first historical vehicle data and the corresponding first historical pure electric range, and the second historical data may include second historical vehicle data and the corresponding second historical pure electric range. Specifically, the first historical pure electric range can be the pure electric range of the first vehicle at the time corresponding to the first historical vehicle data, and the second historical pure electric range can be the pure electric range of the first vehicle at the time corresponding to the second historical vehicle data. The first historical vehicle data may include historical vehicle driving data, historical vehicle environmental data, and historical battery state data; the second vehicle data may include the historical SOC, historical maximum battery pack temperature, and historical minimum battery pack temperature of the second vehicle. The battery may include multiple cells, and the temperatures of different cells may differ. The historical maximum battery pack temperature may be the temperature of the cell with the highest temperature among the multiple cells of the battery, and the historical minimum battery pack temperature may be the temperature of the cell with the lowest temperature among the multiple cells of the battery.

[0038] Specifically, the means of obtaining the first historical data and the second historical data are not limited here. Any means that can obtain the first historical data and the second historical data are within the protection scope of the embodiments of this application.

[0039] Regarding S120, since the mapping relationship between the vehicle's pure electric range and vehicle data differs between the high SOC range (including fully charged state) and the low SOC range, different range prediction models can be used to predict the pure electric range when the vehicle is in the high SOC range and the low SOC range. Therefore, it is necessary to train different range prediction models separately for the high SOC range and the low SOC range.

[0040] Specifically, a first range prediction model can be trained based on multiple first training samples, and a second range prediction model can be trained based on multiple second training samples. The first range prediction model can be used to predict the pure electric range of the target vehicle when it is in a high SOC range, and the second range prediction model can be used to predict the pure electric range of the target vehicle when it is in a low SOC range. Correspondingly, the first historical data in the first training samples can be data from when the first vehicle is in a high SOC range, and the second historical data in the second training samples can be data from when the second vehicle is in a low SOC range. SOC can be understood as the remaining battery charge, usually expressed as a percentage, with a maximum value of 100% and a minimum value of 0%. A high SOC range can be defined as an SOC less than or equal to 100% and greater than or equal to a first threshold (e.g., 30% ≤ SOC ≤ 100%), and a low SOC range can be defined as an SOC greater than or equal to 0% and less than the first threshold (e.g., 0% ≤ SOC < 30%). Since the maximum value of SOC is 100% and the minimum value is 0%, the high SOC range can be restricted to only SOC greater than or equal to the first threshold, and the low SOC range can be restricted to only SOC less than the first threshold.

[0041] In some implementations, to improve the accuracy of the first range prediction model, the first range prediction model is trained based on multiple first training samples until it converges, resulting in the trained first range prediction model. This may include:

[0042] The first training sample is input into the first range prediction model, and the first range prediction model is used to predict the pure electric range of the first vehicle to obtain the first predicted pure electric range.

[0043] Based on the first historical pure electric range and the first predicted pure electric range, the model parameters of the first range prediction model are adjusted until the first range prediction model converges, thus obtaining the trained first range prediction model.

[0044] Here, the first training sample is input into the preset first range prediction model to obtain the first predicted pure electric range corresponding to the first historical vehicle data in the first training sample. Then, based on the first predicted pure electric range and the first historical pure electric range corresponding to the first training sample, the loss function value of the first range prediction model is determined. If the loss function value does not meet the training stopping condition, the model parameters of the first range prediction model are adjusted, and the first range prediction model with adjusted parameters is trained using the first training sample until the training stopping condition is met, thus obtaining the trained first range prediction model.

[0045] This can make the accuracy of the first driving range prediction model, which has been trained, higher.

[0046] In some implementations, to improve the accuracy of the second range prediction model, the second range prediction model is trained based on multiple second training samples until it converges, resulting in the trained second range prediction model. This may include:

[0047] The second training sample is input into the second range prediction model, and the second range prediction model is used to predict the pure electric range of the second vehicle to obtain the second predicted pure electric range.

[0048] Based on the second historical pure electric range and the second predicted pure electric range, the model parameters of the second range prediction model are adjusted until the second range prediction model converges, thus obtaining the trained second range prediction model.

[0049] Here, the second training sample is input into the preset second range prediction model to obtain the second predicted pure electric range corresponding to the second historical vehicle data in the second training sample. Then, based on the second predicted pure electric range and the second historical pure electric range corresponding to the second training sample, the loss function value of the second range prediction model is determined. If the loss function value does not meet the training stopping condition, the model parameters of the second range prediction model are adjusted, and the second range prediction model with adjusted parameters is trained using the second training sample until the training stopping condition is met, thus obtaining the trained second range prediction model.

[0050] This can make the accuracy of the trained second range prediction model higher.

[0051] In some implementations, in order to make the pure electric range predicted by the first range prediction model and the second range prediction model closer to the actual value, historical vehicle driving data may include at least one of historical energy consumption per kilometer, historical acceleration, and historical average vehicle speed.

[0052] Historical vehicle environmental data may include historical outdoor temperatures;

[0053] Historical battery status data may include at least one of the following: historical battery health status (SOH), historical battery pack maximum temperature, historical battery pack minimum temperature, historical state of charge (SOC), historical battery voltage, and historical average battery current.

[0054] Here, the historical battery state data can be the historical state data of the battery that powers the first vehicle. The battery can include multiple cells, and the temperature of different cells can be different. The highest historical battery pack temperature can be the temperature of the cell with the highest temperature among the multiple cells of the battery, and the lowest historical battery pack temperature can be the temperature of the cell with the lowest temperature among the multiple cells of the battery.

[0055] It should be noted that the data required to predict the pure electric range may differ for different vehicle models. Therefore, the specific data types included in the first and second historical vehicle data can be adjusted according to the actual vehicle model, and are not limited here.

[0056] Thus, by using real data such as historical vehicle driving data, historical vehicle environmental data, and historical battery state data, along with the actual pure electric range at the corresponding time, a first range prediction model is trained. By using real data such as historical SOC, historical battery pack maximum temperature, and historical battery pack minimum temperature, along with the actual pure electric range at the corresponding time, a second range prediction model is trained. This allows the pure electric range predicted by the trained first and second range prediction models to be closer to the actual value.

[0057] In some implementations, in order to more accurately predict the pure electric range of the vehicle, the first range prediction model can be the XGBoost model; the second range prediction model can be the Linear Regression model.

[0058] Here, since the mapping relationship between the first historical vehicle data and the first historical pure electric range is closer to the XGBoost model in the high SOC range, the pure electric range of the vehicle in the high SOC range can be predicted by the XGBoost model.

[0059] However, the XGBoost model requires a large amount of historical data as training samples for training. Most users rarely use the remaining battery power of their vehicles down to 0% SOC. Therefore, there is very little historical data in the low SOC range. If the pure electric range in the low SOC range is still trained and predicted based on the XGBoost model, the regression learning ability of the pure electric range will be poor due to the small number of training samples, resulting in unsatisfactory prediction accuracy.

[0060] Since there is a strong linear relationship between the second historical vehicle data and the second historical pure electric range in the low SOC range, the pure electric range of a vehicle in the low SOC range can be predicted using the Linear Regression model.

[0061] In some examples, tests showed that when the SOC was 0% ≤ SOC ≤ 100%, the mean squared error of predicting the pure electric range of the vehicle using the XGBoost model was 7.66 km, while when the SOC was 30% ≤ SOC ≤ 100%, the mean squared error of predicting the pure electric range of the vehicle using the XGBoost model was 7.38 km, and when the SOC was 0% ≤ SOC < 30%, the mean squared error of predicting the pure electric range of the vehicle using the Linear Regression model was 7.38 km.

[0062] When the SOC is between 0% and 30%, the mean square error of the pure electric range predicted by the XGBoost model is 4.07 km, and when the SOC is between 0% and 30%, the mean square error of the pure electric range predicted by the Linear Regression model is 2.43 km.

[0063] Therefore, based on this test, it can be seen that using the Linear Regression model to predict the pure electric range of a vehicle is more accurate in the low SOC range.

[0064] Therefore, using the XGBoost model to predict the pure electric range of a vehicle in the high SOC range and the Linear Regression model to predict the pure electric range in the low SOC range can improve the accuracy of pure electric range prediction.

[0065] The following is in conjunction with the appendix Figure 2 The driving range prediction method provided in the embodiments of this application will be described in detail.

[0066] Figure 2 The diagram illustrates a flowchart of a range prediction method provided in an embodiment of this application. The main body executing this range prediction method can be a range prediction device. Figure 2 As shown, the driving range prediction method provided in this application embodiment may include S210-S230:

[0067] S210, Obtain the battery state of charge (SOC) of the target vehicle;

[0068] S220, when the SOC of the target vehicle is greater than or equal to the first threshold, the first vehicle data of the target vehicle is input into the first range prediction model, and the pure electric range of the target vehicle is predicted by the first range prediction model to obtain the first pure electric range.

[0069] S230: When the SOC of the target vehicle is less than the first threshold, the second vehicle data of the target vehicle is input into the second range prediction model, and the pure electric range of the target vehicle is predicted by the second range prediction model to obtain the second pure electric range.

[0070] Therefore, when the target vehicle's State of Charge (SOC) is greater than or equal to a first threshold, a first range prediction model can be used to predict the target vehicle's pure electric range; when the target vehicle's SOC is less than the first threshold, a second range prediction model can be used to predict the target vehicle's pure electric range. This allows for more accurate predictions of the vehicle's pure electric range through machine learning models. Furthermore, the first range prediction model is trained using first historical vehicle data from a first vehicle of the same model as the target vehicle, obtained when the first vehicle's SOC is greater than or equal to the first threshold. The second range prediction model is trained using second historical vehicle data from a second vehicle of the same model as the target vehicle, obtained when the second vehicle's SOC is less than the first threshold. Using different range prediction models for vehicles with different SOCs further improves the accuracy of pure electric range prediction.

[0071] Regarding S210, since the mapping relationship between the pure electric range of the target vehicle and the vehicle data is different in the high SOC range (including the fully charged state) and the low SOC range, different range prediction models can be used to predict the pure electric range of the target vehicle in the high SOC range and the low SOC range, which requires obtaining the SOC of the target vehicle first.

[0072] Regarding S220, if the State of Charge (SOC) of the target vehicle is greater than or equal to a first threshold, i.e., the target vehicle is in the high SOC range, a first range prediction model can be used to predict the pure electric range of the target vehicle, thus obtaining the first pure electric range. The first vehicle data may include vehicle driving data, vehicle environmental data, and battery state data. The first range prediction model can be trained based on first historical vehicle data of a first vehicle of the same model as the target vehicle. The first historical vehicle data can be obtained when the SOC of the first vehicle is greater than or equal to the first threshold. The first vehicle data may include vehicle driving data, vehicle environmental data, and battery state data.

[0073] Regarding S230, if the target vehicle's State of Charge (SOC) is less than the first threshold, meaning the target vehicle is in a low SOC range, a second range prediction model can be used to predict the target vehicle's pure electric range, thus obtaining the second pure electric range. The second vehicle data may include the target vehicle's SOC, the highest battery pack temperature, and the lowest battery pack temperature. The second range prediction model can be trained based on second historical vehicle data from a second vehicle of the same model as the target vehicle. This second historical vehicle data can be obtained when the second vehicle's SOC is less than the first threshold. The second vehicle data may include the target vehicle's SOC, the highest battery pack temperature, and the lowest battery pack temperature.

[0074] In some implementations, in order to make the predicted pure electric range closer to the actual value, vehicle driving data may include at least one of energy consumption per kilometer, acceleration, and average vehicle speed.

[0075] Vehicle environmental data may include outdoor temperature;

[0076] Battery status data may include at least one of the following: battery health status (SOH), maximum battery pack temperature, minimum battery pack temperature, state of charge (SOC), battery voltage, and average battery current.

[0077] The meaning of the first vehicle data and its various data is the same as that of the first historical vehicle data and its various historical data, except that the first vehicle data and its various data are real-time data, while the first historical vehicle data and its various historical data are historical data, which will not be elaborated further here.

[0078] Thus, predicting the pure electric range in the high SOC range based on the current first vehicle data and predicting the pure electric range in the low SOC range based on the current second vehicle data can make the predicted pure electric range closer to the actual value.

[0079] In some implementations, in order to more accurately predict the pure electric range of the vehicle, the first range prediction model can be the XGBoost model; the second range prediction model can be the Linear Regression model.

[0080] Therefore, using the XGBoost model to predict the pure electric range of a vehicle in the high SOC range and the Linear Regression model to predict the pure electric range in the low SOC range can improve the accuracy of pure electric range prediction.

[0081] Based on the same inventive concept, this application also provides a driving range prediction device. The following describes... Figure 3 The driving range prediction device provided in the embodiments of this application will be described in detail.

[0082] Figure 3 A schematic diagram of the structure of a range prediction device provided in one embodiment of this application is shown.

[0083] like Figure 3 As shown, the range prediction device may include:

[0084] The first acquisition module 301 is used to acquire the battery state of charge (SOC) of the target vehicle.

[0085] The first prediction module 302 is used to input the first vehicle data of the target vehicle into the first range prediction model when the SOC of the target vehicle is greater than or equal to the first threshold, and use the first range prediction model to predict the pure electric range of the target vehicle to obtain the first pure electric range. The first vehicle data may include vehicle driving data, vehicle environment data and battery status data. The first range prediction model may be trained based on the first historical vehicle data of the first vehicle with the same model as the target vehicle. The first historical vehicle data may be obtained when the SOC of the first vehicle is greater than or equal to the first threshold.

[0086] The second prediction module 303 is used to input the second vehicle data of the target vehicle into the second range prediction model when the SOC of the target vehicle is less than the first threshold, and use the second range prediction model to predict the pure electric range of the target vehicle to obtain the second pure electric range. The second vehicle data may include the SOC of the target vehicle, the maximum temperature of the battery pack, and the minimum temperature of the battery pack. The second range prediction model may be trained based on the second historical vehicle data of the second vehicle with the same model as the target vehicle. The second historical vehicle data may be obtained when the SOC of the second vehicle is less than the first threshold.

[0087] Therefore, when the target vehicle's State of Charge (SOC) is greater than or equal to a first threshold, a first range prediction model can be used to predict the target vehicle's pure electric range; when the target vehicle's SOC is less than the first threshold, a second range prediction model can be used to predict the target vehicle's pure electric range. This allows for more accurate predictions of the vehicle's pure electric range through machine learning models. Furthermore, the first range prediction model is trained using first historical vehicle data from a first vehicle of the same model as the target vehicle, obtained when the first vehicle's SOC is greater than or equal to the first threshold. The second range prediction model is trained using second historical vehicle data from a second vehicle of the same model as the target vehicle, obtained when the second vehicle's SOC is less than the first threshold. Using different range prediction models for vehicles with different SOCs further improves the accuracy of pure electric range prediction.

[0088] In some implementations, in order to make the predicted pure electric range closer to the actual value, vehicle driving data may include at least one of energy consumption per kilometer, acceleration, and average vehicle speed.

[0089] Vehicle environmental data may include outdoor temperature;

[0090] Battery status data may include at least one of the following: battery health status (SOH), maximum battery pack temperature, minimum battery pack temperature, state of charge (SOC), battery voltage, and average battery current.

[0091] In some implementations, in order to more accurately predict the pure electric range of the vehicle, the first range prediction model can be the XGBoost model; the second range prediction model can be the Linear Regression model.

[0092] In some implementations, to obtain a more accurate first and second range prediction model, the first and second range prediction models can be trained using the following apparatus:

[0093] The second acquisition module is used to acquire multiple first training samples and multiple second training samples. The first training samples include first historical data of a first vehicle with the same model as the target vehicle, and the second training samples include second historical data of a second vehicle with the same model as the target vehicle.

[0094] The model training module is used to train a first range prediction model based on multiple first training samples until the first range prediction model converges, thus obtaining the trained first range prediction model; and to train a second range prediction model based on multiple second training samples until the second range prediction model converges, thus obtaining the trained second range prediction model.

[0095] In some implementations, to improve the accuracy of the first range prediction model, the model training module may include:

[0096] The first prediction submodule is used to input the first training sample into the first range prediction model, and use the first range prediction model to predict the pure electric range of the first vehicle to obtain the first predicted pure electric range. The first historical data includes the first historical vehicle data and the first historical pure electric range corresponding to the first historical vehicle data.

[0097] The first parameter adjustment submodule is used to adjust the model parameters of the first range prediction model based on the first historical pure electric range and the first predicted pure electric range, until the first range prediction model converges, thus obtaining the trained first range prediction model.

[0098] In some implementations, to improve the accuracy of the second range prediction model, the model training module may include:

[0099] The second prediction submodule is used to input the second training sample into the second range prediction model, and use the second range prediction model to predict the pure electric range of the second vehicle to obtain the second predicted pure electric range. The second historical data includes the second historical vehicle data and the second historical pure electric range corresponding to the second historical vehicle data.

[0100] The second parameter adjustment submodule is used to adjust the model parameters of the second range prediction model based on the second historical pure electric range and the second predicted pure electric range, until the second range prediction model converges, thus obtaining the trained second range prediction model.

[0101] Figure 4 A schematic diagram of the structure of an electronic device provided in one embodiment of this application is shown.

[0102] like Figure 4 As shown, the electronic device 4 is a structural diagram of an exemplary hardware architecture of an electronic device capable of implementing the range prediction method and range prediction device according to the embodiments of this application. This electronic device may refer to the electronic device in the embodiments of this application.

[0103] The electronic device 4 may include a processor 401 and a memory 402 storing computer program instructions.

[0104] Specifically, the processor 401 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0105] Memory 402 may include a large-capacity memory for data or instructions. For example, and not limitingly, memory 402 may include a hard disk drive (HDD), a floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 402 may include removable or non-removable (or fixed) media. Where appropriate, memory 402 may be internal or external to an integrated gateway disaster recovery device. In a particular embodiment, memory 402 is non-volatile solid-state memory. In a particular embodiment, memory 402 may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, electrical, optical, or other physical / tangible memory storage devices. Thus, generally, memory 402 includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this application.

[0106] The processor 401 reads and executes computer program instructions stored in the memory 402 to implement any of the driving range prediction methods in the above embodiments.

[0107] In one example, the electronic device may also include a communication interface 403 and a bus 404. Wherein, as... Figure 4 As shown, the processor 401, memory 402, and communication interface 403 are connected through bus 404 and complete communication with each other.

[0108] The communication interface 403 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0109] Bus 404 includes hardware, software, or both, that couples components of an electronic device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 404 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0110] The electronic device can execute the range prediction method in the embodiments of this application, thereby achieving a combination of Figures 1 to 3 The described method and apparatus for predicting driving range.

[0111] Furthermore, in conjunction with the range prediction methods in the above embodiments, this application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the range prediction methods in the above embodiments.

[0112] Furthermore, in conjunction with the range prediction method in the above embodiments, this invention can provide a vehicle for implementation. The vehicle includes the range prediction device or range prediction equipment described in the above embodiments.

[0113] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0114] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0115] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0116] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, 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, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0117] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for predicting driving range, characterized in that, include: Obtain the battery state of charge (SOC) of the target vehicle; When the SOC of the target vehicle is greater than or equal to a first threshold, the first vehicle data of the target vehicle is input into a first range prediction model. The first range prediction model is used to predict the pure electric range of the target vehicle to obtain a first pure electric range. The first vehicle data includes vehicle driving data, vehicle environmental data, and battery status data. The first range prediction model is trained based on the first historical vehicle data of a first vehicle of the same model as the target vehicle. The first historical vehicle data is obtained when the SOC of the first vehicle is greater than or equal to the first threshold. The first range prediction model is an XGBoost model. When the SOC of the target vehicle is less than the first threshold, the second vehicle data of the target vehicle is input into the second range prediction model. The second range prediction model is used to predict the pure electric range of the target vehicle to obtain the second pure electric range. The second vehicle data includes the SOC of the target vehicle, the highest temperature of the battery pack, and the lowest temperature of the battery pack. The second range prediction model is trained based on the second historical vehicle data of a second vehicle of the same model as the target vehicle. The second historical vehicle data is obtained when the SOC of the second vehicle is less than the first threshold. The second range prediction model is a linear regression model. The battery includes multiple cells. The highest temperature of the battery pack is the temperature of the cell with the highest temperature among the multiple cells of the battery, and the lowest temperature of the battery pack is the temperature of the cell with the lowest temperature among the multiple cells of the battery.

2. The method as described in claim 1, characterized in that, The vehicle driving data includes at least one of energy consumption per kilometer, acceleration, and average vehicle speed; The vehicle environmental data includes outdoor temperature; The battery status data includes at least one of the following: battery health status (SOH), maximum battery pack temperature, minimum battery pack temperature, state of charge (SOC) of the target vehicle, battery voltage, and average battery current.

3. The method as described in claim 1, characterized in that, The first driving range prediction model and the second driving range prediction model were trained using the following method: Acquire multiple first training samples and multiple second training samples, wherein the first training samples include first historical data of a first vehicle that is the same model as the target vehicle, and the second training samples include second historical data of a second vehicle that is the same model as the target vehicle; The first driving range prediction model is trained based on multiple first training samples until the first driving range prediction model converges, resulting in a trained first driving range prediction model. The second driving range prediction model is trained based on multiple second training samples until the second driving range prediction model converges, resulting in a trained second driving range prediction model.

4. The method as described in claim 3, characterized in that, The step of training the first driving range prediction model based on multiple first training samples until the first driving range prediction model converges, to obtain the trained first driving range prediction model, includes: The first training sample is input into the first range prediction model, and the pure electric range of the first vehicle is predicted using the first range prediction model to obtain the first predicted pure electric range. The first historical data includes the first historical vehicle data and the first historical pure electric range corresponding to the first historical vehicle data. Based on the first historical pure electric range and the first predicted pure electric range, the model parameters of the first range prediction model are adjusted until the first range prediction model converges, thus obtaining the trained first range prediction model.

5. The method as described in claim 3, characterized in that, The step of training the second driving range prediction model based on multiple second training samples until the second driving range prediction model converges, to obtain the trained second driving range prediction model, includes: The second training sample is input into the second range prediction model, and the pure electric range of the second vehicle is predicted using the second range prediction model to obtain the second predicted pure electric range. The second historical data includes the second historical vehicle data and the second historical pure electric range corresponding to the second historical vehicle data. Based on the second historical pure electric range and the second predicted pure electric range, adjust the model parameters of the second range prediction model until the second range prediction model converges to obtain the trained second range prediction model.

6. A driving range prediction device, characterized in that, The device includes: The first acquisition module is used to acquire the battery state of charge (SOC) of the target vehicle. A first prediction module is used to input first vehicle data of the target vehicle into a first range prediction model when the SOC of the target vehicle is greater than or equal to a first threshold, and use the first range prediction model to predict the pure electric range of the target vehicle to obtain a first pure electric range. The first vehicle data includes vehicle driving data, vehicle environmental data, and battery status data. The first range prediction model is trained based on first historical vehicle data of a first vehicle of the same model as the target vehicle. The first historical vehicle data is obtained when the SOC of the first vehicle is greater than or equal to the first threshold. The first range prediction model is an XGBoost model. The second prediction module is used to input the second vehicle data of the target vehicle into the second range prediction model when the SOC of the target vehicle is less than the first threshold, and use the second range prediction model to predict the pure electric range of the target vehicle to obtain the second pure electric range. The second vehicle data includes the SOC of the target vehicle, the highest temperature of the battery pack, and the lowest temperature of the battery pack. The second range prediction model is trained based on the second historical vehicle data of a second vehicle of the same model as the target vehicle. The second historical vehicle data is obtained when the SOC of the second vehicle is less than the first threshold. The second range prediction model is a linear regression model. The battery includes multiple cells, the highest temperature of the battery pack is the temperature of the cell with the highest temperature among the multiple cells of the battery, and the lowest temperature of the battery pack is the temperature of the cell with the lowest temperature among the multiple cells of the battery.

7. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the driving range prediction method as described in any one of claims 1-5.

8. A computer storage medium, characterized in that, The computer storage medium stores computer program instructions, which, when executed by a processor, implement the driving range prediction method as described in any one of claims 1-5.

9. A vehicle, characterized in that, The vehicle includes: the range prediction device as claimed in claim 6 or the electronic device as claimed in claim 7.

Images