Vehicle energy consumption estimation method, vehicle, equipment and program product

Abstract

The invention relates to the technical field of vehicles, in particular to a vehicle energy consumption estimation method, a vehicle, equipment and a program product. The method applied to the cloud comprises the steps of obtaining a target model corresponding to a target vehicle; the estimated power consumption predicted by the energy consumption prediction model corresponding to the target model is obtained, and the actual power consumption of the same type of vehicles corresponding to the target model in the actual travel is collected; determining an estimated accurate index value based on the estimated power consumption and the actual power consumption; if the estimated accurate index value reaches the index threshold value, the energy consumption estimation model is updated to the target vehicle, and the target vehicle obtains the estimated power consumption of the vehicle based on the energy consumption estimation model. The accuracy of the estimated power consumption of the target vehicle can be improved, and the driving safety and the use experience of the target vehicle are improved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicles, and particularly to a vehicle energy consumption prediction method, a vehicle, a device and a program product. Background Art

[0002] With the strong support of national policies and the development of the vehicle networking business, electric vehicles are becoming more and more popular, and map route planning has been widely applied to electric vehicles. For most electric vehicles, the driving range is relatively limited. Especially in the case of long-distance and high-speed driving, how to quickly find a suitable charging station has become a matter of great concern to users. However, the reasonable recommendation of charging stations is inseparable from the accurate prediction of energy consumption. In the prior art, based on the energy consumption during future driving, the remaining driving range of the vehicle is determined. Then, by obtaining the geographical location information and the remaining driving range of the target vehicle, when the remaining driving range is less than a preset threshold, the available driving mileage is calculated, and the optimal charging station is selected from multiple charging stations and the location information of the optimal charging station is displayed to the user for the user to choose. However, when the energy consumption prediction is not accurate enough, there is a risk that the remaining energy of the vehicle cannot support the vehicle to reach the recommended charging station, or there is a risk that the remaining energy is too much when the vehicle reaches the recommended charging station, resulting in an increase in the number of vehicle charging times. Furthermore, there are risks of lower driving safety and lower user experience of the vehicle. Summary of the Invention

[0003] Based on the above defects and deficiencies of the prior art, the present application provides a vehicle energy consumption prediction method, a vehicle, a device and a program product, which can improve the accuracy of predicting the power consumption of the target vehicle itself and enhance the driving safety and user experience of the target vehicle.

[0004] According to a first aspect of the present application, there is provided a vehicle energy consumption prediction method, which is applied to the cloud and includes: obtaining a target model corresponding to a target vehicle; obtaining a predicted power consumption predicted by an energy consumption prediction model corresponding to the target model, and collecting an actual power consumption of a same-type vehicle corresponding to the target model during an actual journey; determining a prediction accuracy index value based on the predicted power consumption and the actual power consumption; if the prediction accuracy index value reaches an index threshold, updating the energy consumption prediction model to the target vehicle, wherein the target vehicle obtains its own predicted power consumption based on the energy consumption prediction model.

[0005] According to the vehicle energy consumption prediction method provided in the first aspect of the present application, the determining the prediction accuracy index value based on the predicted power consumption and the actual power consumption includes: if, during the collection process of the actual power consumption, the real-time number of vehicles of the same-type vehicle reaches a vehicle number threshold, and the actual sampling number of the actual power consumption reaches a number threshold, then determining the prediction accuracy index value based on the predicted power consumption and the actual power consumption.

[0006] According to the vehicle energy consumption prediction method provided in the first aspect of the present application, the prediction accuracy index value includes the coefficient of determination; determining the prediction accuracy index value based on the predicted power consumption and the actual power consumption includes: calculating the sum of squared residuals of the predicted power consumption and the actual power consumption, and the total sum of squares of the predicted power consumption and the actual power consumption; calculating the ratio of the sum of squared residuals to the total sum of squares; and calculating the coefficient of determination based on the ratio.

[0007] According to the vehicle energy consumption prediction method provided in the first aspect of the present application, the energy consumption prediction model is trained from the sample data corresponding to the target model, and the sample data includes sample power consumption and sample power consumption factor data that cause the sample power consumption; the training process of the energy consumption prediction model is as follows: obtaining historical driving data generated by at least one vehicle during a historical journey, where the historical driving data is data generated by the vehicle during the historical journey; based on the historical driving data, obtaining the sample power consumption and the sample power consumption factor data corresponding to each vehicle respectively; and training an original neural network model based on the sample power consumption and the sample power consumption factor data of each vehicle corresponding to the target model to obtain the energy consumption prediction model.

[0008] According to the vehicle energy consumption prediction method provided in the first aspect of the present application, the energy consumption prediction model includes a sub-energy consumption prediction model; training the original neural network model based on the sample power consumption and the sample power consumption factor data of each vehicle corresponding to the target model to obtain the energy consumption prediction model includes: classifying the sample power consumption and the sample power consumption factor data according to at least one preset mileage range based on the mileage of the historical journey corresponding to the target model to obtain sub-sample data corresponding to each preset mileage range, where the sub-sample data includes the sample power consumption and the sample power consumption factor data corresponding to the preset mileage range; and training the original neural network model with each type of sub-sample data to obtain the sub-energy consumption prediction models corresponding to each preset mileage range respectively.

[0009] According to the vehicle energy consumption prediction method provided in the first aspect of the present application, after collecting the actual power consumption of the same-type vehicles corresponding to the target model during an actual journey, it further includes: updating the actual power consumption to the sample power consumption, and updating the power consumption factor data corresponding to the actual journey to the sample power consumption factor data; after updating the energy consumption prediction model to the target vehicle, it further includes: retraining the original neural network model based on the updated sample power consumption and the sample power consumption factor to obtain a retrained energy consumption prediction model.

[0010] According to the vehicle energy consumption prediction method provided in the first aspect of the present application, obtaining the predicted power consumption obtained by predicting with the energy consumption prediction model corresponding to the target model includes: determining whether there is sample data corresponding to the target model. If so, training the energy consumption prediction model based on the sample data corresponding to the target model; if not, determining a reference vehicle with the highest similarity in vehicle parameters to the target vehicle, and using the reference energy consumption prediction model configured for the reference vehicle as the energy consumption prediction model; obtaining the predicted power consumption based on the energy consumption prediction model.

[0011] According to the second aspect of the present application, a vehicle is provided. The vehicle is used to receive the energy consumption prediction model transmitted from the cloud, and is used to obtain the predicted power consumption of the vehicle itself based on the energy consumption prediction model; wherein, the cloud implements the vehicle energy consumption prediction method according to any one of the first aspect.

[0012] According to the third aspect of the present application, an electronic device is provided, including: a memory and a processor; the memory is connected to the processor and is used to store a program; the processor is used to implement the vehicle energy consumption prediction method according to the first aspect by running the program in the memory.

[0013] According to the fourth aspect of the present application, a computer program product is provided, including computer program instructions;

[0014] When the computer program instructions are run by a processor, the processor is caused to execute the vehicle energy consumption prediction method according to the first aspect.

[0015] In the present application, the target model corresponding to the target vehicle is obtained; the predicted power consumption obtained by predicting with the energy consumption prediction model corresponding to the target model is obtained, and the actual power consumption of the same-type vehicles corresponding to the target model during the actual journey is collected; based on the predicted power consumption and the actual power consumption, a predicted accuracy index value is determined; if the predicted accuracy index value reaches the index threshold, the energy consumption prediction model is updated to the target vehicle, where the target vehicle obtains the predicted power consumption of the vehicle itself based on the energy consumption prediction model. In the above solution, the predicted power consumption of the target vehicle predicted by the energy consumption prediction model and the actual power consumption of the same-type vehicles corresponding to the target model are used to evaluate the prediction accuracy of the energy consumption prediction model. If the predicted accuracy index value reaches the index threshold, it indicates that the prediction accuracy of the energy consumption prediction model is relatively high. At this time, updating the energy consumption prediction model to the target vehicle can ensure that the predicted power consumption of the vehicle itself obtained by the target vehicle based on the energy consumption prediction model is more accurate, thereby improving the driving safety and user experience of the target vehicle. Description of the Drawings

[0016] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only the embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on the provided drawings.

[0017] Figure 1 A flowchart of a vehicle energy consumption prediction method provided by an embodiment of the present application;

[0018] Figure 2 A block diagram of a vehicle energy consumption prediction device provided by an embodiment of the present application;

[0019] Figure 3 A schematic structural diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0021] Exemplary method

[0022] To improve the accuracy of vehicle energy consumption prediction, the present application provides a vehicle energy consumption prediction method. This method is implemented in the cloud independent of the target vehicle, thereby reducing the computing pressure on the local of the target vehicle. The cloud is a device with computing and processing capabilities and can perform data interaction with the target vehicle and other vehicles. For example, a remote server, etc.

[0023] In one embodiment, as Figure 1 shown, the implementation process of the vehicle energy consumption prediction method applied to the cloud is as follows:

[0024] Step 101, obtain the target model corresponding to the target vehicle.

[0025] In this embodiment, the target vehicle is a vehicle powered by electric energy. The target vehicle can perform data interaction with the cloud. When the target vehicle is started, the target model can be transmitted to the cloud based on manual information or automatic information, and the cloud obtains the target model corresponding to the target vehicle.

[0026] Step 102, obtain the predicted power consumption predicted by the energy consumption prediction model corresponding to the target model, and collect the actual power consumption of the same-type vehicles corresponding to the target model during the actual journey.

[0027] In this embodiment, each vehicle model corresponds to an energy consumption prediction model, which is trained using the sample data under the vehicle model. Therefore, the energy consumption prediction model is more suitable for the vehicles under the vehicle model. Vehicles of the same vehicle model use the energy consumption prediction model under the vehicle model, and the energy consumption prediction is more accurate. After the cloud obtains the target model, the energy consumption prediction model corresponding to the target model is used for energy consumption prediction to obtain the predicted power consumption.

[0028] In this embodiment, in order to further improve the accuracy of energy consumption prediction, the energy consumption prediction model is not directly configured in the target vehicle. Instead, the accuracy of the energy consumption prediction model is evaluated to determine whether the prediction accuracy of the energy consumption prediction model under the target model is high enough. When evaluating the energy consumption prediction model, in addition to the predicted power consumption obtained by using the energy consumption prediction model, the actual power consumption also needs to be collected. The actual power consumption includes not only the actual power consumption during the actual journey of the target vehicle, but also the actual power consumption during the actual journey of other vehicles of the same type under the target model. Based on multiple vehicles of the same type, the risk of accidental deviation caused by collecting the actual power consumption of a single vehicle can be avoided.

[0029] In this embodiment, when using the energy consumption prediction model for energy consumption prediction, the data input into the energy consumption prediction model can be the driving data when the vehicles of the same type under the target model generate the actual power consumption. That is to say, the predicted power consumption and the actual power consumption are the predicted value and the actual value under the same driving conditions. In this way, the evaluation accuracy of the energy consumption prediction model can be further improved.

[0030] Step 103: Determine the predicted accuracy index value based on the predicted power consumption and the actual power consumption.

[0031] In this embodiment, after obtaining the predicted power consumption and the actual power consumption, the predicted accuracy index value is determined based on the predicted power consumption and the actual power consumption. The predicted accuracy index value is used to evaluate the prediction accuracy of the energy consumption prediction model. Specifically, the predicted accuracy index value can be set according to the actual situation. For example, the average error percentage of the predicted power consumption and the actual power consumption is determined as the predicted accuracy index value, or the coefficient of determination is determined as the predicted accuracy index value.

[0032] Step 104: If the predicted accuracy index value reaches the index threshold, update the energy consumption prediction model to the target vehicle, where the target vehicle obtains the predicted power consumption of its own vehicle based on the energy consumption prediction model.

[0033] In this embodiment, an index threshold is set in advance according to the actual situation and requirements. If the estimated accurate index value reaches the index threshold, it indicates that the estimation accuracy of the energy consumption estimation model has reached the expectation, and then the energy consumption estimation model can be updated to the target vehicle. During the use of the target vehicle, the energy consumption estimation model can be used to more accurately predict the estimated power consumption of the vehicle itself, so as to perform processes such as remaining mileage prediction and charging pile recommendation based on the estimated power consumption of the vehicle itself. If the estimated accurate index value does not reach the index threshold, it indicates that the estimation accuracy of the energy consumption estimation model does not reach the expectation, and then the energy consumption estimation model can be optimized and trained again until the estimation accuracy of the energy consumption estimation model reaches the expectation, and then the energy consumption estimation model is updated to the target vehicle. The target vehicle has higher accuracy in predicting the estimated power consumption using the energy consumption estimation model, which can further ensure higher driving safety and better user experience of the target vehicle.

[0034] In one embodiment, based on the estimated power consumption and the actual power consumption, the estimated accurate index value is determined, including: if during the actual power consumption collection process, the real-time number of vehicles of the same type reaches the vehicle number threshold, and the actual sampling number of the actual power consumption reaches the number threshold, then based on the estimated power consumption and the actual power consumption, the estimated accurate index value is determined.

[0035] In this embodiment, the real-time number of vehicles and the actual sampling number are used to determine the timing of determining the estimated accurate index value, so as to realize the automatic update of the energy consumption estimation model and improve the degree of automation. Specifically, the cloud receives a large amount of driving data of many vehicles of various models, so as to update the model for various vehicle models. For the target model among various models, the real-time number of vehicles of the same type that upload form data to the cloud under the target model is counted. When the real-time number of vehicles reaches the vehicle number threshold, it indicates that the real-time number of vehicles under the target model has accumulated to a certain quantity. And for the same vehicle under the target model, it may have actually driven multiple times, and each actual drive will generate corresponding actual power consumption. When the actual sampling number reaches the number threshold, it indicates that the actual power consumption sampling number of multiple vehicles of the same type has accumulated to a certain quantity. When the real-time number of vehicles of the same type reaches the vehicle number threshold and the actual sampling number of the actual power consumption reaches the number threshold, the timing of determining the estimated accurate index value is reached, that is, the update timing of the energy consumption estimation model, thus providing conditions for realizing the automatic update of the energy consumption estimation model.

[0036] In this embodiment, the vehicle number threshold and the number threshold are set according to the actual situation and requirements. For example, the vehicle number threshold is set to 100 and the number threshold is set to 1000.

[0037] In one embodiment, the estimated accuracy index value includes the coefficient of determination. Based on the estimated power consumption and the actual power consumption, the estimated accuracy index value is determined, including: calculating the sum of squared residuals of the estimated power consumption and the actual power consumption, and the total sum of squares of the estimated power consumption and the actual power consumption; calculating the ratio of the sum of squared residuals to the total sum of squares; and calculating the coefficient of determination based on the ratio.

[0038] In this embodiment, the estimated accuracy index value includes the coefficient of determination (also known as R 2 ), and the process of determining the estimated accuracy index value based on the estimated power consumption and the actual power consumption is the process of calculating the coefficient of determination based on the estimated power consumption and the actual power consumption. The formula for the coefficient of determination R 2 is as follows:

[0039]

[0040] The sum of squared residuals (RSS) is the sum of the squares of the differences between the estimated power consumption and the actual power consumption, which reflects the data variability that the energy consumption prediction model fails to explain. The total sum of squares (TSS) is the sum of the squares of the differences between the actual power consumption and the average of the estimated power consumption, which represents the total variability of the data itself.

[0041] In this embodiment, when the sum of squared residuals is smaller, that is, the estimated power consumption output by the energy consumption prediction model is closer to the actual power consumption, the value of R 2 will be larger, indicating that the fitting effect of the energy consumption prediction model is better. When the sum of squared residuals is close to the total sum of squares, that is, the energy consumption prediction model can hardly explain the data variability, the value of R 2 will be close to 0, indicating that the fitting effect of the model is poor. If the sum of squared residuals is equal to 0, it means that the energy consumption prediction model perfectly fits the data, and at this time R 2 is equal to 1. If the sum of squared residuals is greater than the total sum of squares, R 2 may become negative, which usually means that the performance of the energy consumption prediction model is worse than a simple mean prediction.

[0042] In this embodiment, using the coefficient of determination can accurately reflect the prediction ability of the energy consumption prediction model, thus ensuring that the updated energy consumption prediction model in the target vehicle has higher prediction accuracy.

[0043] In this embodiment, the coefficient of determination is calculated based on the actual power consumption generated during each actual trip of each homogeneous vehicle and the estimated power consumption of the homogeneous vehicle. Each coefficient of determination is compared with a determination threshold respectively. The number of coefficients of determination exceeding the determination threshold is counted. If the ratio of the number of coefficients of determination exceeding the determination threshold to the total number of coefficients of determination reaches a ratio threshold, it indicates that the prediction ability of the energy consumption prediction model is good enough, and then the energy consumption prediction model is automatically updated to the target vehicle. The determination threshold and the ratio threshold can be set according to the actual situation and needs. For example, the determination threshold is set to 0.9 and the ratio threshold is set to 0.9.

[0044] In one embodiment, the estimated accuracy index value further includes the average error percentage between the estimated power consumption and the actual power consumption. The average error percentage is calculated based on the actual power consumption generated during each actual trip of each homogeneous vehicle and the estimated power consumption of the homogeneous vehicle respectively. The smaller the average error percentage, the better the prediction ability of the energy consumption prediction model. The value of the average error percentage can be observed manually to determine whether to update the energy consumption prediction model to the target vehicle. The average error percentage can also be compared with a percentage threshold. If the average error percentage reaches the percentage threshold, the energy consumption prediction model can be automatically updated to the target vehicle.

[0045] In one embodiment, the energy consumption prediction model is trained from the sample data corresponding to the target model. The sample data includes sample power consumption and sample power consumption factor data that cause the sample power consumption.

[0046] The training process of the energy consumption prediction model is as follows: Obtain the historical driving data generated by at least one vehicle during historical trips, where the historical driving data is the data generated by the vehicle during historical trips; Based on the historical driving data, obtain the sample power consumption and sample power consumption factor data corresponding to each vehicle respectively; Based on the sample power consumption and sample power consumption factor data of each vehicle corresponding to the target model, train the original neural network model to obtain the energy consumption prediction model.

[0047] In this embodiment, the cloud collects the driving data of numerous vehicles in real time. The driving data includes parameters such as the vehicle model, vehicle identification number (VIN), vehicle load, motor model of the vehicle, and data such as driving duration, driving speed, battery power, vehicle interior and exterior temperature generated in real time during the vehicle driving process. Any vehicle reports the driving data to the cloud during the driving process, and the cloud receives the driving data reported by each vehicle and stores it. After a vehicle completes a driving process, this driving process becomes a historical trip, and the data generated during this historical trip is historical driving data.

[0048] In this embodiment, the cloud references databases such as StarRocks that can store a large amount of data. StarRocks is a high-performance real-time analysis database mainly used to process complex queries and analysis tasks on large-scale data sets. The data in the database is stored indexed by date and vehicle model, facilitating the storage, retrieval, and processing of historical driving data.

[0049] In this embodiment, the vehicle continuously sends historical driving data to the cloud, and the cloud continuously updates the historical driving data in the database. Further, the cloud periodically cleans the data in the database to obtain and continuously update the sample data used for training the energy consumption prediction model. Optionally, based on the vehicle identification number of each vehicle, the historical power consumption in each historical trip is extracted from the historical driving data of each vehicle as the sample power consumption, and the data of the power consumption factors related to each historical power consumption is extracted as the sample power consumption factor data. Specifically, the power consumption factors that cause the vehicle to generate historical power consumption include, but are not limited to, road conditions, temperature difference between the inside and outside of the vehicle, vehicle speed, etc. Multiple power consumption factors can more comprehensively reflect the power consumption of the vehicle. Further, there are various interference factors in vehicle energy consumption. For example, when the vehicle is in standby state and the air conditioner is running but the vehicle speed is zero, although there is energy consumption, there is no mileage; traffic jams; extremely high-speed driving; interference factors will cause abnormal power consumption of the vehicle. If the data caused by interference factors is directly involved in training and fed back to the energy consumption prediction model, it will cause the energy consumption prediction model to deviate from the facts and reduce the accuracy. To avoid noise interference, the initially extracted sample power consumption and sample power consumption factor data can be denoised according to a pre-set denoising strategy. For example, the denoising strategy includes deleting the sample power consumption and sample power consumption factor data corresponding to the state where the vehicle speed is zero or the vehicle speed is lower than the speed threshold in the standby state of the vehicle, and also includes deleting the sample power consumption and sample power consumption factor data corresponding to the temperature difference between the inside and outside of the vehicle exceeding 30 degrees. The denoising strategy is set according to the actual use situation and specific requirements of the vehicle.

[0050] In this embodiment, for each vehicle model, the cloud obtains the sample data under each vehicle model through processes such as cleaning, extraction, and denoising. Then, the sample data under each vehicle model is used to train the energy consumption prediction model for that vehicle model respectively. Specifically, based on the sample power consumption and sample power consumption factor data of the vehicles corresponding to the target model, the original neural network model is trained to obtain the energy consumption prediction model.

[0051] In this embodiment, the original neural network model can select the model network architecture and model category according to the actual situation and requirements. For example, the original neural network model can adopt any one of the model architectures such as Multilayer Perceptron (MLP), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN), Hybrid Models, Graph Neural Networks (GNN), and other model architectures. Which model architecture to choose depends on specific requirements, data characteristics, and computing resources in the cloud.

[0052] In one embodiment, the energy consumption prediction model includes sub-energy consumption prediction models.

[0053] Training the original neural network model based on the sample power consumption and sample power consumption factor data of each vehicle corresponding to the target model to obtain the energy consumption prediction model, including: classifying the sample power consumption and sample power consumption factor data according to the travel mileage of the historical trips corresponding to the target model in at least one preset mileage range to obtain sub-sample data respectively corresponding to each preset mileage range, where the sub-sample data includes the sample power consumption and sample power consumption factor data corresponding to the preset mileage range; using various types of sub-sample data to train the original neural network model respectively to obtain sub-energy consumption prediction models respectively corresponding to each preset mileage range.

[0054] In this embodiment, through analysis, when the driving mileage of a vehicle journey is different, the power consumption of the vehicle journey shows different patterns. Therefore, based on a preset mileage range, the sample data can be classified, and then the sub-sample data of each category can be used to train the sub-energy consumption prediction models respectively corresponding to each preset mileage range. When the predicted accuracy index value reaches the index threshold, each sub-energy consumption prediction model corresponding to a preset mileage range is updated to the target vehicle. When the target vehicle applies each sub-energy consumption prediction model, it can determine the mileage of the upcoming journey in any way such as through navigation or manual input before the journey starts, and then call the sub-energy consumption prediction model corresponding to the preset mileage range where the mileage is located, further improving the prediction accuracy of the energy consumption prediction model. The preset mileage range is preset according to the actual situation and needs. For example, the preset mileage range includes 0 - 20 km, 20 - 50 km, 50 - 100 km, and above 100 km. The sample data is divided into 4 categories of sub-sample data according to the journey mileage of each historical journey, that is, the sub-sample data corresponding to 0 - 20 km, 20 - 50 km, 50 - 100 km, and above 100 km respectively. The 4 categories of sub-sample data are used to train the original neural network model respectively, and the sub-energy consumption prediction models corresponding to 0 - 20 km, 20 - 50 km, 50 - 100 km, and above 100 km are obtained.

[0055] In this embodiment, the model architectures of each sub-energy consumption prediction model are the same, but due to the different sub-sample data used, the model parameters of the sub-energy consumption prediction models are different, thus ensuring that each sub-energy consumption prediction model is more suitable for each preset mileage range.

[0056] In one embodiment, after collecting the actual power consumption of the same-type vehicles corresponding to the target model in the actual journey, it further includes: updating the actual power consumption to the sample power consumption, and updating the power consumption factor data corresponding to the actual journey to the sample power consumption factor data. After updating the energy consumption prediction model to the target vehicle, it further includes: re-training the original neural network model based on the updated sample power consumption and sample power consumption factors to obtain the re-trained energy consumption prediction model.

[0057] In this embodiment, each vehicle including the target vehicle continuously uploads driving data to the cloud during the vehicle journey. The cloud can not only obtain the actual power consumption from the real-time uploaded driving data, but also update the actual power consumption to the sample power consumption, and update the power consumption factor data corresponding to the actual journey to the sample power consumption factor data, continuously increasing the number of sample data. Taking the target vehicle as an example, after updating the energy consumption prediction model to the target vehicle, the updated sample power consumption and sample power consumption factors are used to retrain the original neural network model or optimize the training of the energy consumption prediction model to obtain the retrained energy consumption prediction model. Specifically, after updating the energy consumption prediction model to the target vehicle, the real-time vehicle number and the actual sampling number are re-counted; while re-counting the real-time vehicle number and the actual sampling number, model training or optimization is performed again based on the updated sample power consumption and sample power consumption factors to obtain the retrained energy consumption prediction model. If the re-counted real-time vehicle number reaches the vehicle number threshold again and the actual sampling number reaches the number threshold again, the retrained energy consumption prediction model is updated to the target vehicle, thereby realizing the continuous optimization of the energy consumption prediction model and the cyclic update of the energy consumption prediction model in the target vehicle.

[0058] In one embodiment, obtaining the predicted power consumption predicted by the energy consumption prediction model corresponding to the target model includes: determining whether there is sample data corresponding to the target model, if so, training the energy consumption prediction model based on the sample data corresponding to the target model; if not, determining the reference vehicle with the highest similarity of vehicle parameters to the target vehicle, and using the reference energy consumption prediction model configured in the reference vehicle as the energy consumption prediction model; obtaining the predicted power consumption based on the energy consumption prediction model.

[0059] In this embodiment, if the cloud has accumulated sample data corresponding to the target model, the energy consumption prediction model is trained using the sample data and is used in the implementation process of this method. If the target model is a newly connected vehicle model and the cloud has not received any driving data of the new vehicle model, the cloud cannot obtain the sample data of the new vehicle model, then the reference vehicle with the highest similarity of vehicle parameters to the target vehicle is determined, and the energy consumption prediction model in the reference vehicle is a model trained with sample data. The reference energy consumption prediction model configured in the reference vehicle is used as the energy consumption prediction model in the implementation process of this method. After temporarily using the reference energy consumption prediction model configured in the reference vehicle as the energy consumption prediction model, the target vehicle and the same-type vehicles under the target model continuously upload driving data to the cloud, and then implement the continuous optimization of the energy consumption prediction model described in the above embodiment and the cyclic update process of the energy consumption prediction model in the target vehicle, continuously improving the prediction accuracy of the energy consumption prediction model.

[0060] In this embodiment, weights can be assigned to vehicle models (such as large vehicles, medium-sized vehicles, sedans, SUVs, multi-purpose vehicles, etc.), the number of drive motors (such as front-wheel drive, four-wheel drive), the mileage at full charge, vehicle weight, and other parameters to obtain the similarity between the target vehicle and each other vehicle model, and then the reference vehicle can be determined. It should be noted that for the vehicle models used to calculate the similarity, the cloud has stored the energy consumption prediction models trained with sample data.

[0061] In one embodiment, after the energy consumption prediction model is trained, the energy consumption prediction model can be first put into the pre-release environment. The energy consumption prediction model is updated to the pre-release vehicles of the target model, and the pre-release vehicles conduct test runs based on the energy consumption prediction model to obtain test run data. Then, the training samples corresponding to the energy consumption prediction model and the test run data obtained from the test runs of the pre-release vehicles are used as test data to calculate the estimated accuracy index value. When the estimated accuracy index value reaches the index threshold, the energy consumption prediction model is really updated to the target vehicle. The pre-release environment test provided by this embodiment is more applicable to the situation where a new vehicle model is initially connected to the cloud, avoiding the situation where the reference energy consumption prediction model has poor adaptability to the target vehicle as the energy consumption prediction model, resulting in potential safety hazards for the target vehicle. Or, it is possible to manually configure whether the update process of the energy consumption prediction model goes through the pre-release environment to improve the flexibility and safety of the update of the energy consumption prediction model.

[0062] In this application, the target model corresponding to the target vehicle is obtained; the estimated power consumption predicted by the energy consumption prediction model corresponding to the target model is obtained, and the actual power consumption of the same-type vehicles corresponding to the target model during the actual journey is collected; based on the estimated power consumption and the actual power consumption, the estimated accuracy index value is determined; if the estimated accuracy index value reaches the index threshold, the energy consumption prediction model is updated to the target vehicle, where the target vehicle obtains the estimated power consumption of this vehicle based on the energy consumption prediction model. In the above solution, the estimated power consumption of the target vehicle predicted by the energy consumption prediction model and the actual power consumption of the same-type vehicles corresponding to the target model are used to evaluate the prediction accuracy of the energy consumption prediction model. If the estimated accuracy index value reaches the index threshold, it indicates that the prediction accuracy of the energy consumption prediction model is relatively high. At this time, updating the energy consumption prediction model to the target vehicle can ensure that the estimated power consumption of this vehicle obtained by the target vehicle based on the energy consumption prediction model is more accurate, thereby improving the driving safety and usage experience of the target vehicle.

[0063] Furthermore, the estimated power consumption of this vehicle obtained by the target vehicle based on the energy consumption prediction model can be applied to various purposes such as charging station recommendation and intelligent driving strategy adjustment. The higher the accuracy of the estimated power consumption of this vehicle, the higher the driving safety of the target vehicle and the better the usage experience.

[0064] Furthermore, the energy consumption prediction model realizes an optimized training and cyclic update self-learning process. The cloud can continuously connect to new vehicle models and complete automated negotiation for new vehicle models, with strong scalability, improving the convenience of later operations and reducing subsequent maintenance costs.

[0065] Exemplary vehicle

[0066] Correspondingly, an embodiment of the present application further provides a vehicle, which is used to receive the energy consumption prediction model transmitted by the cloud and to obtain the predicted power consumption of the vehicle based on the energy consumption prediction model. Among them, the cloud implements the vehicle energy consumption prediction method provided in any one of the above embodiments.

[0067] In one embodiment, the vehicle is used to upload the driving data during the driving process to the cloud.

[0068] In one embodiment, the vehicle is used to receive at least one sub-energy consumption prediction model. Before the vehicle starts a journey, based on any one of the methods such as navigation or manual input, the mileage of the upcoming journey is determined, and then the sub-energy consumption prediction model corresponding to the preset mileage range where the mileage is located is called.

[0069] The vehicle provided in this embodiment belongs to the same inventive concept as the vehicle energy consumption prediction method provided in the above embodiments of the present application. The vehicle energy consumption prediction method provided in any one of the above embodiments of the present application can be applied, and it has the corresponding functional modules and beneficial effects for implementing the method. For the technical details not described in detail in this embodiment, reference can be made to the specific processing content of the vehicle energy consumption prediction method provided in the above embodiments of the present application, which will not be elaborated here.

[0070] Exemplary device

[0071] Correspondingly, an embodiment of the present application further provides a cloud device, as Figure 2 shown. This device may include:

[0072] A model type acquisition module 201, configured to acquire the target model corresponding to the target vehicle;

[0073] A power consumption acquisition module 202, configured to acquire the predicted power consumption predicted by the energy consumption prediction model corresponding to the target model, and collect the actual power consumption of the same-type vehicles corresponding to the target model during the actual journey;

[0074] An index determination module 203, configured to determine the predicted accuracy index value based on the predicted power consumption and the actual power consumption;

[0075] A model update module 204, configured to update the energy consumption prediction model to the target vehicle if the predicted accuracy index value reaches the index threshold, where the target vehicle obtains the predicted power consumption of the vehicle based on the energy consumption prediction model.

[0076] In one embodiment, the index determination module 203 is configured to determine a predicted accurate index value based on the predicted power consumption and the actual power consumption if, during the process of collecting the actual power consumption, the real-time number of vehicles of the same type reaches the vehicle number threshold and the actual number of samples of the actual power consumption reaches the number threshold.

[0077] In one embodiment, the predicted accurate index value includes the coefficient of determination;

[0078] The index determination module 203 is configured to calculate the sum of squared residuals of the predicted power consumption and the actual power consumption, and the total sum of squares of the predicted power consumption and the actual power consumption; calculate the ratio of the sum of squared residuals to the total sum of squares; and calculate the coefficient of determination based on the ratio.

[0079] In one embodiment, the energy consumption prediction model is trained by using sample data corresponding to the target model, and the sample data includes sample power consumption and sample power consumption factor data that cause the sample power consumption;

[0080] The cloud device further includes a model training module for training the energy consumption prediction model, and the training process is as follows: obtaining historical driving data generated by at least one vehicle during a historical journey, where the historical driving data is data generated by the vehicle during the historical journey; based on the historical driving data, obtaining the sample power consumption and the sample power consumption factor data respectively corresponding to each vehicle; and training the original neural network model based on the sample power consumption and the sample power consumption factor data of each vehicle corresponding to the target model to obtain the energy consumption prediction model.

[0081] In one embodiment, the energy consumption prediction model includes a sub-energy consumption prediction model;

[0082] The training module is configured to classify the sample power consumption and the sample power consumption factor data according to at least one preset mileage range based on the mileage of the historical journey corresponding to the target model, to obtain sub-sample data respectively corresponding to each preset mileage range, where the sub-sample data includes the sample power consumption and the sample power consumption factor data corresponding to the preset mileage range; and use each type of sub-sample data to train the original neural network model respectively to obtain sub-energy consumption prediction models respectively corresponding to each preset mileage range.

[0083] In one embodiment, the training module is further configured to update the actual power consumption to the sample power consumption, and update the power consumption factor data corresponding to the actual journey to the sample power consumption factor data; and re-train the original neural network model based on the updated sample power consumption and the sample power consumption factor to obtain the re-trained energy consumption prediction model.

[0084] In one embodiment, the power consumption acquisition module 202 is configured to determine whether there is sample data corresponding to the target model. If so, an energy consumption prediction model is trained based on the sample data corresponding to the target model; if not, a reference vehicle with the highest similarity to the vehicle parameters of the target vehicle is determined, and the reference energy consumption prediction model configured for the reference vehicle is used as the energy consumption prediction model; the predicted power consumption is obtained based on the energy consumption prediction model.

[0085] The vehicle energy consumption prediction device provided in this embodiment belongs to the same inventive concept as the vehicle energy consumption prediction method provided in the above embodiments of the present application, and can execute the vehicle energy consumption prediction method provided in any of the above embodiments of the present application, and has the corresponding functional modules and beneficial effects for executing the method. For the technical details not described in detail in this embodiment, reference may be made to the specific processing content of the vehicle energy consumption prediction method provided in the above embodiments of the present application, which will not be elaborated here.

[0086] Exemplary electronic device

[0087] An embodiment of the present application also provides an electronic device, as Figure 3 shown, the electronic device includes: a memory 300 and a processor 301.

[0088] The memory 300 is connected to the processor 301 and is used to store programs.

[0089] The processor 301 is configured to implement the vehicle energy consumption prediction method in the above embodiment by running the program stored in the memory 300.

[0090] Specifically, the above electronic device may further include: a communication interface 302, an input device 303, an output device 304, and a bus 305.

[0091] The processor 301, the memory 300, the communication interface 302, the input device 303, and the output device 304 are interconnected through the bus. Among them:

[0092] The bus 305 may include a path for transmitting information between various components of the computer system.

[0093] The processor 301 may be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present invention. It may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0094] The processor 301 may include a main processor, and may also include a baseband chip, a modem, etc.

[0095] The memory 300 stores a program for implementing the technical solution of the present invention, and may also store an operating system and other key services. Specifically, the program may include program code, and the program code includes computer operation instructions. More specifically, the memory 300 may include a read-only memory (ROM), other types of static storage devices that can store static information and instructions, a random access memory (RAM), other types of dynamic storage devices that can store information and instructions, a disk memory, a flash memory, etc.

[0096] The input device 303 may include a device for receiving data and information input by a user, such as a keyboard, a mouse, a camera, a scanner, a light pen, a voice input device, a touch screen, a pedometer, or a gravity sensor, etc.

[0097] The output device 304 may include a device for allowing information to be output to a user, such as a display screen, a printer, a speaker, etc.

[0098] The communication interface 302 may include a device of any transceiver type for communicating with other devices or communication networks, such as an Ethernet, a radio access network (RAN), a wireless local area network (WLAN), etc.

[0099] The processor 301 executes the program stored in the memory 300 and calls other devices, and can be used to implement each step of the vehicle energy consumption prediction method provided in the above embodiments of the present application.

[0100] Exemplary computer program product and storage medium

[0101] In addition to the above methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, and the computer program instructions, when run by a processor, cause the processor to execute the steps in the vehicle energy consumption prediction method described in the embodiments of the present application.

[0102] The computer program product may be written in any combination of one or more programming languages for programming code for performing the operations of the embodiments of the present application. The programming languages include object-oriented programming languages, such as Java, C++, etc., and also include conventional procedural programming languages, such as the "C" language or similar programming languages. The program code may be executed entirely on a user computing device, partially on a user device, executed as an independent software package, partially on a user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0103] In addition, an embodiment of the present application may also be a storage medium, on which a computer program is stored, and the computer program is executed by a processor to perform the steps in the vehicle energy consumption prediction method described in the embodiments of the present application.

[0104] For the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present application is not limited by the described action sequence, because according to the present application, certain steps may be in other sequences or performed simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present application.

[0105] It should be noted that the embodiments in this specification are all described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0106] The steps in the methods of the embodiments of the present application can be adjusted, combined, and deleted according to actual needs, and the technical features recorded in each embodiment can be replaced or combined.

[0107] The modules and sub-modules in the devices and terminals provided in the embodiments of the present application can be combined, divided, and deleted according to actual needs.

[0108] In several embodiments provided by the present application, it should be understood that the disclosed terminals, devices, and methods can be implemented in other ways. For example, the terminal embodiments described above are merely illustrative. For example, the division of modules or sub-modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple sub-modules or modules can be combined or integrated into another module, or some features can be ignored or not executed. Another point, the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of devices or modules can be in an electrical, mechanical, or other form.

[0109] The modules or sub-modules described as separate components may or may not be physically separated. The components as modules or sub-modules may or may not be physical modules or sub-modules, that is, they can be located in one place, or they can be distributed to multiple network modules or sub-modules. Some or all of the modules or sub-modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0110] In addition, in each embodiment of the present application, each functional module or sub-module can be integrated into a processing module, or each module or sub-module can exist physically alone, or two or more modules or sub-modules can be integrated into one module. The above-mentioned integrated module or sub-module can be implemented in the form of hardware, or in the form of a software functional module or sub-module.

[0111] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this document can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Skilled professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0112] The steps of the method or algorithm described in combination with the embodiments disclosed in this document can be directly implemented by hardware, a software unit executed by a processor, or a combination of the two. The software unit can be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium well-known in the technical field.

[0113] Finally, it should also be noted that in this document, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the existence of additional identical elements in the process, method, article or device comprising the element.

[0114] The foregoing description of the disclosed embodiments enables those skilled in the art to practice or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A vehicle energy consumption prediction method, characterized in that, Applied to the cloud, including: Obtain the target model corresponding to the target vehicle; Obtain the estimated power consumption predicted by the energy consumption prediction model corresponding to the target model, and collect the actual power consumption of the same-type vehicles corresponding to the target model during the actual journey; Determine the estimated accuracy index value based on the estimated power consumption and the actual power consumption; If the estimated accuracy index value reaches the index threshold, update the energy consumption prediction model to the target vehicle, where the target vehicle obtains the estimated power consumption of its own vehicle based on the energy consumption prediction model.

2. The vehicle energy consumption prediction method according to claim 1, wherein The determining the estimated accuracy index value based on the estimated power consumption and the actual power consumption includes: If during the process of collecting the actual power consumption, the number of real-time vehicles of the same-type vehicle reaches the vehicle number threshold, and the actual sampling number of the actual power consumption reaches the number threshold, then determine the estimated accuracy index value based on the estimated power consumption and the actual power consumption.

3. The vehicle energy consumption prediction method according to claim 2, wherein The estimated accuracy index value includes the coefficient of determination; The determining the estimated accuracy index value based on the estimated power consumption and the actual power consumption includes: Calculate the sum of squared residuals of the estimated power consumption and the actual power consumption, and the total sum of squares of the estimated power consumption and the actual power consumption; Calculate the ratio of the sum of squared residuals to the total sum of squares; Calculate the coefficient of determination based on the ratio.

4. The vehicle energy consumption prediction method according to claim 1, wherein The energy consumption prediction model is trained by the sample data corresponding to the target model, and the sample data includes sample power consumption and sample power consumption factor data that cause the sample power consumption; The training process of the energy consumption prediction model is as follows: Obtain the historical driving data generated by at least one vehicle during the historical journey, where the historical driving data is the data generated by the vehicle during the historical journey; Based on the historical driving data, obtain the sample power consumption and the sample power consumption factor data corresponding to each of the vehicles; Train the original neural network model based on the sample power consumption and the sample power consumption factor data of each vehicle corresponding to the target model to obtain the energy consumption prediction model.

5. The vehicle energy consumption prediction method according to claim 4, wherein The energy consumption prediction model includes a sub-energy consumption prediction model; The training the original neural network model based on the sample power consumption and the sample power consumption factor data of each vehicle corresponding to the target model to obtain the energy consumption prediction model includes: Classify the sample power consumption and the sample power consumption factor data according to at least one preset mileage range based on the mileage of the historical journey corresponding to the target model to obtain sub-sample data corresponding to each preset mileage range, where the sub-sample data includes the sample power consumption and the sample power consumption factor data corresponding to the preset mileage range; Use each type of sub-sample data to train the original neural network model respectively to obtain the sub-energy consumption prediction models corresponding to each preset mileage range.

6. The vehicle energy consumption estimation method according to claim 4, wherein After collecting the actual power consumption of the same-type vehicles corresponding to the target model, it further includes: Update the actual power consumption to the sample power consumption, and update the power consumption factor data corresponding to the actual driving range to the sample power consumption factor data; After updating the energy consumption prediction model to the target vehicle, it further includes: Based on the updated sample power consumption and the sample power consumption factors, retrain the original neural network model to obtain a retrained energy consumption prediction model.

7. The vehicle energy consumption prediction method according to claim 1, characterized in that, The obtaining of the predicted power consumption predicted by the energy consumption prediction model corresponding to the target model includes: Determine whether there is sample data corresponding to the target model. If so, train the energy consumption prediction model based on the sample data corresponding to the target model; if not, determine the reference vehicle with the highest similarity of vehicle parameters to the target vehicle, and use the reference energy consumption prediction model configured for the reference vehicle as the energy consumption prediction model; Obtain the predicted power consumption based on the energy consumption prediction model.

8. A vehicle, characterized in that, The vehicle is used to receive the energy consumption prediction model transmitted from the cloud, and is used to obtain the predicted power consumption of the vehicle itself based on the energy consumption prediction model; Wherein, the cloud implements the vehicle energy consumption prediction method according to any one of claims 1-7.

9. An electronic device, characterized in that, It includes: A memory and a processor; The memory is connected to the processor and is used to store programs; The processor is used to implement the vehicle energy consumption prediction method according to any one of claims 1-7 by running the programs in the memory.

10. A computer program product, characterized in that, It includes computer program instructions; When the computer program instructions are run by the processor, the processor is caused to execute the vehicle energy consumption prediction method according to any one of claims 1-7.

Images