Energy management method and device for extended-range hybrid electric vehicle

Abstract

The invention relates to the technical field of energy management, in particular to an energy management method and device for extended-range hybrid automobiles, and the method comprises the steps: extracting static features according to the operation information and environment information of a plurality of extended-range hybrid automobiles at at least one historical moment, time sequence features are extracted according to the road condition information and the predicted average vehicle speed of the multiple extended-range hybrid vehicles on the target driving road section after the at least one historical moment, and a training data set is constructed according to the static features, the time sequence features and the actual energy output information; generating a supervised learning label of a multi-generation initial energy management model based on the training data set; and using the training data set, the supervised learning label, the target loss function and the target optimizer to perform end-to-end training on the initial energy management model so as to construct a target energy management model. Therefore, the problems that an energy management method in the prior art has certain limitation under complex and changeable road conditions, the information transmission efficiency is low, optimization is difficult to integrate and the like are solved.

Description

Technical Field

[0001] The present application relates to the field of energy management technology, and in particular to an energy management method and device for an extended-range hybrid vehicle. Background Art

[0002] In the related technologies, the energy management strategies of extended-range hybrid vehicles mostly adopt rule-based or model prediction methods. These methods usually implement the energy consumption prediction model and the energy management control strategy separately, that is, first predict the energy consumption of the extended-range hybrid vehicle, and then manage the energy of the extended-range hybrid vehicle on this basis.

[0003] However, in related technologies, rule-based methods are difficult to adapt to dynamic environments, and the control accuracy based on model prediction is limited by the quality of the model. They both have certain limitations under complex and changeable road conditions. In addition, the separate implementation of the energy consumption prediction model and the energy management control strategy can easily lead to low information transmission efficiency, and the whole process optimization is difficult to integrate, which needs to be solved urgently. Summary of the invention

[0004] The present application provides an energy management method and device for an extended-range hybrid vehicle to solve the problems in related technologies, namely, that the rule-based method is difficult to adapt to dynamic environments, and the control accuracy based on model prediction is limited by the model quality, and both have certain limitations under complex and changeable road conditions. In addition, the energy consumption prediction model and the energy management control strategy are implemented separately, which easily leads to low information transmission efficiency and difficulty in integrating the whole process optimization.

[0005] In a first aspect, an embodiment of the present application provides an energy management method for an extended-range hybrid vehicle, which is applied to a model building stage and includes the following steps: collecting operation information, environmental information and actual energy output information of multiple extended-range hybrid vehicles at at least one historical moment, as well as road condition information and predicted average vehicle speed of a target driving section of the multiple extended-range hybrid vehicles after the at least one historical moment; extracting static features corresponding to the multiple extended-range hybrid vehicles at the at least one historical moment according to the operation information and the environmental information, extracting time series features corresponding to the target driving section of the multiple extended-range hybrid vehicles after the at least one historical moment according to the road condition information and the predicted average vehicle speed, and constructing a training data set according to the static features, the time series features and the actual energy output information; based on the training data set, generating motor torque distribution results, engine torque distribution results and engine start-stop control results of the multiple extended-range hybrid vehicles at the at least one historical moment to generate supervised learning labels for an initial energy management model; using the training data set, the supervised learning labels, the target loss function and the target optimizer, end-to-end training the initial energy management model to construct a target energy management model for managing the energy of the target extended-range hybrid vehicle.

[0006] Through the above technical means, the embodiments of the present application can end-to-end train and optimize the initial energy management model by using a large amount of data sets, supervised learning labels, a target loss function, and a target optimizer to obtain a target energy management model. The target energy management model can directly input information and output the optimal power source torque distribution and engine start-stop control, eliminating the complex modular design and information transmission process in the traditional method, reducing the errors and computational burden in the intermediate links, greatly improving the efficiency of energy management, and contributing to the realization of the whole process optimization.

[0007] Optionally, in an embodiment of the present application, before end-to-end training the initial energy management model by using the training data set, the supervised learning label, the target loss function, and the target optimizer, it further includes: determining energy management optimization objectives corresponding to the plurality of range-extended hybrid electric vehicles based on the energy management requirements corresponding to the plurality of range-extended hybrid electric vehicles; obtaining an optimization result according to the energy management optimization objective, and determining the target loss function according to the optimization result.

[0008] Through the above technical means, the embodiments of the present application can determine the energy management optimization objective according to the energy management requirements of the vehicle, and then determine the target loss function of the model. Thus, the objective of energy management can be transformed into a quantifiable and optimizable mathematical form, providing a clear optimization direction for model training, enabling the model to optimize energy management by minimizing the loss function, and improving performance indicators such as energy utilization efficiency.

[0009] Optionally, in an embodiment of the present application, before end-to-end training the initial energy management model by using the training data set, the supervised learning label, the target loss function, and the target optimizer, it further includes: constructing an energy consumption loss function, a power battery balance loss function, and a smoothness loss function of the target loss function; determining the target loss function based on the energy consumption loss function, the power battery balance loss function, and the smoothness loss function.

[0010] Through the above technical means, the embodiments of the present application can generate a target loss function by comprehensively considering various factors, which can take into account different-dimensional objectives at the same time. For example, while ensuring the minimum energy consumption of the vehicle, it can also ensure the smoothness of driving. Moreover, it helps different loss functions to constrain the target energy management model from different angles, reducing the risk of overfitting, making the target energy management model perform more stably on complex and variable data, and enhancing the robustness and generalization ability of the target energy management model.

[0011] Optionally, in an embodiment of the present application, it further includes: optimizing a preset initial energy management model by using a learning rate decay strategy and an early stopping mechanism to construct the initial energy management model.

[0012] Through the above technical means, the embodiments of the present application can introduce a learning rate decay strategy and an early stopping mechanism to optimize the preset initial energy management model, avoid overfitting of the initial energy management model, and ensure that the obtained target energy management model after training can adapt to new and unseen test data, thereby ensuring the performance of the target energy management model in actual application scenarios.

[0013] An embodiment of the second aspect of the present application provides an energy management method for a range-extended hybrid vehicle, which is applied to the model application stage. Using the above energy management method for a range-extended hybrid vehicle, the method includes the following steps: collecting real-time operation information and real-time environment information of a target range-extended hybrid vehicle, and collecting road condition information and predicted average vehicle speed of the target range-extended hybrid vehicle on a future target driving section; inputting the real-time operation information, the real-time environment information, the road condition information, and the predicted average vehicle speed into a pre-constructed target energy management model, extracting real-time static features of the target range-extended hybrid vehicle and sequential features of the target range-extended hybrid vehicle on the future target driving section, and fusing the static features and the sequential features to obtain a fused feature, so as to output a real-time motor torque distribution result, a real-time engine torque distribution result, and a real-time engine start-stop control result of the target range-extended hybrid vehicle according to the fused feature; generating corresponding control instructions according to the real-time motor torque distribution result, the real-time engine torque distribution result, and the real-time engine start-stop control result, so as to control the target range-extended hybrid vehicle to perform energy management.

[0014] Through the above technical means, the embodiments of the present application can use the trained target energy management model to directly achieve optimal power source torque distribution and engine start-stop control from information input to output, eliminate the complex modular design and information transmission process in the traditional method, reduce the errors and computational burden in the intermediate links, greatly improve the efficiency of energy management, and contribute to realizing the whole-process optimization.

[0015] An embodiment of the third aspect of the present application provides an energy management device for a range-extended hybrid vehicle, which is applied to the model construction stage and includes: a collection module, configured to collect the operation information, environment information, and actual energy output information of multiple range-extended hybrid vehicles at at least one historical moment, and the road condition information and predicted average speed of the target driving section of the multiple range-extended hybrid vehicles after the at least one historical moment; an extraction module, configured to extract the static features corresponding to the multiple range-extended hybrid vehicles at the at least one historical moment according to the operation information and the environment information, extract the time series features corresponding to the target driving section of the multiple range-extended hybrid vehicles after the at least one historical moment according to the road condition information and the predicted average speed, and construct a training data set according to the static features, the time series features, and the actual energy output information; a generation module, configured to generate the motor torque distribution result, engine torque distribution result, and engine start-stop control result of the multiple range-extended hybrid vehicles at the at least one historical moment based on the training data set, so as to generate the supervised learning labels of the initial energy management model; a first construction module, configured to end-to-end train the initial energy management model by using the training data set, the supervised learning labels, the target loss function, and the target optimizer, so as to construct a target energy management model for managing the energy of the target range-extended hybrid vehicle.

[0016] By the above technical means, the embodiment of the present application can end-to-end train and optimize the initial energy management model by using a large amount of data sets, supervised learning labels, a target loss function, and a target optimizer to obtain a target energy management model. The target energy management model can directly perform the optimal power source torque distribution and engine start-stop control from information input to output, eliminating the complex modular design and information transmission process in the traditional method, reducing the errors and computational burden in the intermediate links, greatly improving the efficiency of energy management, and helping to achieve the whole process optimization.

[0017] Optionally, in an embodiment of the present application, it further includes: a first determination module, configured to determine the energy management optimization target of the multiple range-extended hybrid vehicles based on the energy management requirements of the multiple range-extended hybrid vehicles before end-to-end training the initial energy management model by using the training data set, the supervised learning labels, the target loss function, and the target optimizer; a second determination module, configured to obtain an optimization result according to the energy management optimization target, and determine the target loss function according to the optimization result.

[0018] Through the above technical means, the embodiments of the present application can determine the energy management optimization goal according to the energy management requirements of the vehicle, and then determine the target loss function of the model. Thus, the goal of energy management can be transformed into a quantifiable and optimizable mathematical form, providing a clear optimization direction for model training, enabling the model to optimize energy management by minimizing the loss function, and improving performance indicators such as energy utilization efficiency.

[0019] Optionally, in an embodiment of the present application, it further includes: a second construction module, configured to construct an energy consumption loss function, a power battery balance loss function, and a smoothness loss function of the target loss function before end-to-end training the initial energy management model by using the training data set, the supervised learning label, the target loss function, and the target optimizer; a third determination module, configured to determine the target loss function based on the energy consumption loss function, the power battery balance loss function, and the smoothness loss function.

[0020] Through the above technical means, the embodiments of the present application can generate a target loss function by comprehensively considering various factors, taking into account different-dimensional goals at the same time, such as ensuring the minimization of vehicle energy consumption while ensuring driving smoothness, and helping different loss functions to constrain the target energy management model from different angles, reducing the risk of overfitting, making the target energy management model perform more stably on complex and variable data, and enhancing the robustness and generalization ability of the target energy management model.

[0021] Optionally, in an embodiment of the present application, it further includes: a third construction module, configured to optimize a preset initial energy management model by using a learning rate decay strategy and an early stopping mechanism to construct the initial energy management model.

[0022] Through the above technical means, the embodiments of the present application can introduce a learning rate decay strategy and an early stopping mechanism to optimize a preset initial energy management model, avoid overfitting of the initial energy management model, and ensure that the obtained target energy management model after training can adapt to new and unseen test data, so as to ensure the performance of the target energy management model in actual application scenarios.

[0023] In the embodiment of the fourth aspect of the present application, an energy management device for a range-extended hybrid vehicle is provided, which is applied to the model application stage. Using the above-mentioned energy management device for a range-extended hybrid vehicle, it includes: a collection module, configured to collect the real-time operation information and real-time environment information of the target range-extended hybrid vehicle, and collect the road condition information and predicted average speed of the target range-extended hybrid vehicle on the future target driving section; a processing module, configured to input the real-time operation information, the real-time environment information, the road condition information, and the predicted average speed into a pre-constructed target energy management model, extract the real-time static characteristics of the target range-extended hybrid vehicle and the temporal characteristics of the target range-extended hybrid vehicle on the future target driving section, and fuse the static characteristics and the temporal characteristics to obtain a fused characteristic, so as to output the real-time motor torque distribution result, the real-time engine torque distribution result, and the real-time engine start-stop control result of the target range-extended hybrid vehicle according to the fused characteristic; a management module, configured to generate corresponding control instructions according to the real-time motor torque distribution result, the real-time engine torque distribution result, and the real-time engine start-stop control result, so as to control the target range-extended hybrid vehicle to perform energy management.

[0024] By the above technical means, the embodiment of the present application can use the trained target energy management model to directly realize the optimal power source torque distribution and engine start-stop control from information input to output, eliminate the complex modular design and information transmission process in the traditional method, reduce the errors and calculation burden in the intermediate links, greatly improve the efficiency of energy management, and contribute to the realization of the whole-process optimization.

[0025] In the embodiment of the fifth aspect of the present application, a vehicle is provided, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the program to implement the energy management method of the range-extended hybrid vehicle as described in the above embodiment.

[0026] In the embodiment of the sixth aspect of the present application, a computer-readable storage medium is provided, and the computer-readable storage medium stores a computer program, and when the program is executed by a processor, it implements the energy management method of the range-extended hybrid vehicle as above.

[0027] In the embodiment of the seventh aspect of the present application, a computer program product is provided, including a computer program, and when the computer program is executed, it is used to implement the energy management method of the range-extended hybrid vehicle as above.

[0028] In the embodiments of the present application, by collecting various current vehicle information, various information of the future target driving section, and predicting the average vehicle speed, the torque distribution result and the engine start-stop control result can be directly generated through the trained target energy management model, and then the corresponding control instructions are generated to control the target range-extended hybrid vehicle to achieve energy management. Thus, the initial energy management model is trained and optimized end-to-end by using a large amount of data sets, supervised learning labels, target loss functions, and target optimizers to obtain the target energy management model, so that the target energy management model in the present application can directly input information to output the optimal power source torque distribution and engine start-stop control, eliminating the complex modular design and information transmission process in the traditional method, reducing the errors and computational burdens in the intermediate links, thereby greatly improving the efficiency of energy management, contributing to the realization of the whole-process optimization. At the same time, while ensuring that the obtained power source torque distribution result is the global optimal energy management strategy, the reasoning ability of the target energy management model in the present application can realize real-time energy management of the target range-extended hybrid vehicle. When facing complex and changeable road conditions and environmental conditions, it can quickly respond and adjust the energy management strategy, with extremely high practicality while ensuring the robustness and generalization of the target energy management model. Thus, the problems in the related art are solved, that is, the rule-based method is difficult to adapt to the dynamic environment, the control accuracy based on model prediction is limited by the model quality, both have certain limitations under complex and changeable road conditions, and the energy consumption prediction model and the energy management control strategy are separately implemented, which is likely to lead to low information transmission efficiency and difficulty in integrating the whole-process optimization, etc.

[0029] Additional aspects and advantages of the present application will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, in which:

[0031] Figure 1 is a flowchart of an end-to-end range-extended hybrid vehicle energy management method according to an embodiment of the present application;

[0032] Figure 2 is a flowchart of an energy management method for a range-extended hybrid vehicle applied to the model construction stage according to an embodiment of the present application;

[0033] Figure 3 is a schematic diagram of a data set construction and annotation framework according to an embodiment of the present application;

[0034] Figure 4 is a schematic diagram of model training and optimization according to an embodiment of the present application;

[0035] Figure 5 Schematic diagram of the structure of an energy management method and device for a range-extended hybrid vehicle applied in the model building stage according to an embodiment of the present application;

[0036] Figure 6 Flowchart of an energy management method for a range-extended hybrid vehicle applied in the model application stage according to an embodiment of the present application;

[0037] Figure 7 Flowchart of the actual application of the model according to an embodiment of the present application;

[0038] Figure 8 Schematic diagram of the structure of an energy management method and device for a range-extended hybrid vehicle applied in the model application stage according to an embodiment of the present application;

[0039] Figure 9 Schematic diagram of the structure of a vehicle according to an embodiment of the present application.

[0040] Reference numerals:

[0041] 10 - Energy management method and device 10 for a range-extended hybrid vehicle applied in the model building stage: 100 - Acquisition module, 200 - Extraction module, 300 - Generation module, and 400 - Construction module; 10 - Energy management method and device 20 for a range-extended hybrid vehicle applied in the model application stage: 500 - Acquisition module, 600 - Processing module, 700 - Management module; 901 - Memory, 902 - Processor, and 903 - Communication interface. Detailed implementation manners

[0042] The embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present application and should not be construed as a limitation of the present application.

[0043] The energy management method and device for a range-extended hybrid vehicle according to an embodiment of the present application will be described below with reference to the accompanying drawings. In view of the related technologies mentioned in the above background art, the rule-based method is difficult to adapt to dynamic environments, and the control accuracy of model prediction is limited by the model quality. Both have certain limitations under complex and variable road conditions. In addition, the separate implementation of the energy consumption prediction model and the energy management control strategy is likely to lead to low information transmission efficiency and difficulty in integrating the whole-process optimization. The present application provides an energy management method for a range-extended hybrid vehicle. In this method, by collecting various current vehicle information, various information of the future target driving section, and predicting the average vehicle speed, the torque distribution result and the engine start-stop control result can be directly generated by a trained target energy management model, and then the corresponding control instructions are generated to control the target range-extended hybrid vehicle to achieve energy management. Thus, by training the target energy management model through an end-to-end method, it is possible to train and optimize the initial energy management model to obtain the target energy management model by using a large amount of data sets, supervised learning labels, a target loss function, and a target optimizer. The target energy management model in the present application can directly input information to output the optimal power source torque distribution and engine start-stop control, eliminating the complex modular design and information transmission process in the traditional method, reducing the errors and computational burdens in the intermediate links, thereby greatly improving the efficiency of energy management, contributing to the realization of the whole-process optimization. At the same time, while ensuring that the obtained power source torque distribution result is a globally optimal energy management strategy, the inference ability of the target energy management model in the present application can achieve real-time energy management of the target range-extended hybrid vehicle. When facing complex and variable road conditions and environmental conditions, it can quickly respond and adjust the energy management strategy, with extremely high practicality while ensuring the robustness and generalization of the target energy management model. Thus, the problems in the related technologies are solved, such as the rule-based method being difficult to adapt to dynamic environments, the control accuracy of model prediction being limited by the model quality, both having certain limitations under complex and variable road conditions, and the separate implementation of the energy consumption prediction model and the energy management control strategy being likely to lead to low information transmission efficiency and difficulty in integrating the whole-process optimization, etc.

[0044] Before explaining the energy management method for the range-extended hybrid vehicle in the embodiment of the present application, the overall architecture of the energy management model involved in the embodiment of the present application will be explained first.

[0045] In recent years, the booming of artificial intelligence technology has brought revolutionary changes to the energy management field, providing the possibility to build a more intelligent and efficient energy management system. With the development of AI technology, deep learning and reinforcement learning have shown great potential in data-driven prediction and control optimization, and can be used to solve complex optimization problems in the energy management of range-extended hybrid vehicles.

[0046] To this end, the present application can implement the energy management of a range-extended hybrid vehicle based on an AI deep learning model, that is, directly output the power source torque distribution and engine start-stop control from the input of the information of the environment and the vehicle state, thereby achieving global optimization and real-time control.

[0047] It should be noted that in the embodiments of the present application, the preset initial energy management model, the initial energy management model, and the target energy management model involved in the following description steps are essentially end-to-end deep learning models, all including but not limited to three parts: a feature extraction module, a feature fusion module, and an output module.

[0048] Among them, the feature extraction module can be used to extract time series features (time series features) and static features (static global features):

[0049] Time series feature extraction: The dynamic information of the future road section of a range-extended hybrid vehicle after a certain moment includes but is not limited to the average vehicle speed v avg , slope grade, road section length L, and congestion level cong. The above information is input into the LSTM network in the form of a time series. Let the input sequence be X seq ={x 1 , x 2 , …, x T}, where the data dimension of each time step is d seq , and can be but is not limited to represented as follows:

[0050] h t =LSTM(h t-1 , x t ; θ LSTM )

[0051] Among them, h t is the hidden state of the t-th time step, containing the dynamic features of the current time step and the information of the past road section; θ LSTM is the parameter of the LSTM network, including the weight matrix and bias.

[0052] Finally, the embodiments of the present application can but are not limited to take the last hidden state of the LSTM as the time series feature output:

[0053]

[0054] Static global feature extraction: The static state of the current vehicle includes the throttle pedal opening, the brake pedal opening, the battery SOC, the engine speed, the motor speed, the actual vehicle speed, and the ambient temperature. The above information is input into the fully connected network in the form of a feature vector X stat , and its global features are extracted. Its output can be but is not limited to represented as:

[0055]

[0056] Among them, W stat and b stat are the weights and biases of the fully connected layer; F stat is the global feature representation after static feature extraction.

[0057] Feature fusion module:

[0058] In the embodiment of the present application, the time series feature F seq and the static feature F stat can but are not limited to be combined through the feature fusion module to generate the global feature representation. By concatenating F seq and F stat and inputting them into a multi-layer fully connected network. The fusion process can but is not limited to be expressed as:

[0059]

[0060] Among them, [F seq , F stat is the feature concatenation; W fusion and b fusion are the weights and biases of the fusion network; F global is the fused global feature representation, serving as the final feature representation of the model.

[0061] Output module:

[0062] In the output module, the embodiment of the present application can but is not limited to generate the allocation strategies of the engine torque T global , the motor torque T engine , and the engine start-stop state State motor respectively through three groups of fully connected networks based on the global feature F engine :

[0063] T engine = W eng F global + b eng

[0064] T motor = W motor F global + b motor

[0065] S engine = σ(W state F global + b state ).

[0066] Among them, W eng , W motor and W stateis the weight matrix of the output layer; b eng , b motor and b state is the bias of the output layer; T engine and T motor respectively represent the torque outputs of the engine and the motor, S engine represents the engine start-stop state, and σ represents the sigmoid activation function.

[0067] By utilizing the inference ability of the deep learning model, the embodiments of the present application can achieve real-time management of vehicle energy. Especially when facing complex and changeable road conditions and environmental conditions, it can quickly respond and adjust strategies, effectively solving the defects such as low transmission efficiency and poor real-time performance in traditional methods.

[0068] Figure 1 is a flowchart of an end-to-end range-extended hybrid vehicle energy management method according to an embodiment of the present application. As Figure 1 shown, the end-to-end range-extended hybrid vehicle energy management method in the embodiments of the present application is divided into a model construction stage and a model application stage.

[0069] Specifically, Figure 2 is a flowchart of an energy management method for a range-extended hybrid vehicle applied to the model construction stage provided by the embodiments of the present application.

[0070] As Figure 2 shown, the energy management method for the range-extended hybrid vehicle applied to the model construction stage includes the following steps:

[0071] In step S201, collect the operation information, environmental information, and actual energy output information of multiple range-extended hybrid vehicles at at least one historical moment, as well as the road condition information and predicted average vehicle speed of the target driving section of multiple range-extended hybrid vehicles at at least one historical moment later.

[0072] In some embodiments, the present application can construct a unified deep learning model, which can directly obtain the optimal power source torque distribution and engine start-stop control from the input of the vehicle future section working condition information and other vehicle information and output them. Therefore, before constructing this model, the embodiments of the present application can first perform a certain design on the information to be input into the model. It should be noted that in the description here and in the subsequent processes of the embodiments of the present application, the vehicle refers to a range-extended hybrid vehicle.

[0073] For example, the embodiments of the present application can collect the operation information and environmental information of multiple range-extended hybrid vehicles at at least one historical moment. It should be noted that the at least one historical moment selected for each range-extended hybrid vehicle among the multiple range-extended hybrid vehicles can be selected or adjusted according to the actual situation, that is, the at least one historical moment selected for each range-extended hybrid vehicle can be the same or different. Here, only an exemplary illustration is given, and no specific limitation is made.

[0074] In the embodiments of the present application, the operation information includes but is not limited to the throttle pedal opening, brake pedal opening, battery SOC, engine speed, motor speed, and actual vehicle speed of the vehicle at a certain historical moment, and the environmental information includes but is not limited to the environmental temperature of the vehicle at a certain historical moment.

[0075] Furthermore, in order to ensure the accuracy of the model output result, the present application can design the input data more comprehensively. Therefore, the embodiments of the present application can also collect the road condition information and predicted average vehicle speed of the target driving section of the vehicle after a certain historical moment. Here, the target driving section can be understood as the future driving section of the vehicle after a certain historical moment. The specific selection rule of the target driving section can be set by those skilled in the art according to the actual situation. For example, it can be set as the driving section within a certain time period (such as 30s, 1min) after a certain historical moment, or the driving section within a certain length (such as 500m, 1km) after a certain historical moment, etc. Only an exemplary illustration is given in the embodiments of the present application, and no specific limitation is made.

[0076] The predicted average vehicle speed here refers to the average vehicle speed of the vehicle predicted in the target driving section at this historical moment. The specific calculation rule can be selected or adjusted by those skilled in the art according to the actual situation of the vehicle or actual needs. For example, it can be calculated according to the vehicle driving situation of the vehicle in a period of time before a certain historical moment (such as the average vehicle speed in the 1min or 1km before a certain historical moment), or calculated according to the average driving vehicle speed of other vehicles in the networking system, or calculated according to the vehicle driving situation of the vehicle in a period of time before a certain historical moment combined with the road condition information of the target driving section (such as calculated according to the average vehicle speed in the 1min or 1km before a certain historical moment combined with the road condition information of the target driving section). Only an exemplary illustration is given in the embodiments of the present application, and no specific limitation is made.

[0077] And, in the embodiments of the present application, the collected road condition information includes but is not limited to information such as the slope, length, and congestion degree of the target driving section.

[0078] Additionally, in order to facilitate the later determination that the model can adapt to the actual driving scenario of the vehicle, the embodiments of the present application can also collect the actual energy output information of the vehicle at a certain historical moment, thereby providing comprehensive input data for the model.

[0079] The embodiments of the present application can collect a large amount of historical data of vehicles to prepare data for the dataset required for model construction, and effectively ensure the effectiveness of model training through a large amount of comprehensive data.

[0080] Step S202: Extract static features corresponding to multiple range-extended hybrid vehicles at at least one historical moment according to the running information and environmental information, extract temporal features corresponding to the target driving section after at least one historical moment for multiple range-extended hybrid vehicles according to the road condition information and predicted average vehicle speed, and construct a training dataset according to the static features, temporal features, and actual energy output information.

[0081] In some other embodiments, the present application can also perform feature extraction on the collected data, which can effectively improve the accuracy and precision of the output results while saving the computing power of the model.

[0082] For example, Figure 3 is a schematic diagram of the dataset construction and annotation framework for an embodiment of the present application. As Figure 3 shown, the embodiments of the present application can extract static features of multiple range-extended hybrid vehicles at at least one historical moment according to the running information and environmental information. The network architecture used for extracting static features can be, but is not limited to, a fully connected network.

[0083] Furthermore, the embodiments of the present application can extract temporal features corresponding to the target driving section after at least one historical moment for multiple range-extended hybrid vehicles according to the road condition information and predicted average vehicle speed. The network architecture used for extracting temporal features can be, but is not limited to, an LSTM (Long Short-Term Memory) network.

[0084] The embodiments of the present application can construct a dataset through the temporal features, static features, and actual energy output information extracted from a large amount of actual vehicle historical data, so that the finally trained model can dynamically adjust the energy management strategy according to different input features and has strong adaptability. Whether in extreme working conditions such as urban congested sections, highways, and steep slopes, the model can reasonably allocate power sources to ensure the stable operation of the system, effectively improving the practicability and application scope of the model in the present application.

[0085] It should be noted that in addition to constructing a training dataset, the embodiments of the present application can also construct a certain validation dataset and test dataset at the same time. For example, 70% of all data is divided into a training dataset, 20% is divided into a validation dataset, and 10% is divided into a test dataset. Among them, the training dataset can be used, but is not limited to, optimizing model parameters, the validation dataset can be used, but is not limited to, adjusting hyperparameters and preventing overfitting, and the test dataset can be used, but is not limited to, evaluating the performance of the trained model on unknown data.

[0086] Step S203: Based on the training data set, generate the motor torque distribution results, engine torque distribution results, and engine start / stop control results of multiple range-extended hybrid vehicles at at least one historical moment, so as to generate the supervised learning labels of the initial energy management model.

[0087] As a possible implementation manner, the learning method used by the model in the embodiments of the present application is supervised learning. In order to provide the optimal labels for the supervised learning of the model, the embodiments of the present application can calculate the optimal power source allocation strategy and engine start / stop control strategy of multiple range-extended hybrid vehicles at at least one historical moment based on the training data set, that is, the optimal motor torque distribution result, the optimal engine torque distribution result, and the engine start / stop control result, and use them as the supervised learning labels of the initial energy management model. Herein, the initial energy management model can be understood as an end-to-end deep learning model for managing vehicle energy that has not been trained and tested.

[0088] For example, still as Figure 3 shown, the present application can, but is not limited to, adopt the Dynamic Programming (DP) algorithm to generate the optimal power source allocation strategy for each sample. Herein, each sample can be understood as a sample of the static features, temporal features, and actual energy output, etc. corresponding to each moment after dividing the road section into multiple discrete moments.

[0089] Through the DP algorithm, the present application can achieve obtaining the strategy that minimizes the vehicle energy consumption through global optimization, and further provide a reliable supervision signal for the deep learning model.

[0090] Assume that the objective function of the system is as follows:

[0091] J = ∫(fuel consumption(T engine ) + λ·|ΔSOC| + μ·ΔT smooth )dt

[0092] wherein, fuel consumption(T engine ) is the engine fuel consumption, ΔSOC is the SOC deviation, which is used to constrain the battery state to avoid overcharging or discharging, and ΔT smooth is the power source switching smoothness index, and λ and μ are weight coefficients.

[0093] For each sample, the embodiments of the present application can search for the optimal power distribution scheme of the engine and the motor under the current entire working condition through dynamic programming. Finally, record the optimal engine torque, motor torque, and engine start / stop state corresponding to each moment as the supervised learning labels.

[0094] The embodiments of the present application can use a dynamic programming algorithm to generate the optimal power source allocation strategy for each sample as the supervised learning label of the model, which can provide optimal labels for the supervised learning of the model, ensure that the model can learn an energy management strategy close to the global optimum during the model training process, and ensure the global optimality of the results output by the final model, without the need to run a complex optimization process in real time.

[0095] Step S204: Use the training data set, supervised learning label, target loss function, and target optimizer to end-to-end train the initial energy management model to construct a target energy management model for managing the energy of the target range-extended hybrid vehicle.

[0096] In some embodiments, after obtaining the training data set, the embodiments of the present application can use the training data set and supervised learning label in combination with the target loss function and target optimizer to end-to-end train the initial energy management model to construct a target energy management model for managing the energy of the target range-extended hybrid vehicle.

[0097] Among them, the target loss function can be understood here as a function that measures the difference between the model output result and the supervised learning label; the target optimizer can be understood here as various stochastic optimization algorithms that can adaptively adjust the learning rate for each training parameter in the target energy management model.

[0098] For example, Figure 4 is a schematic diagram of model training and optimization for an embodiment of the present application. As Figure 4 shown, the present application can input the training data set and supervised learning label into the initial energy management model, and then perform end-to-end training using the Adam optimizer, and design a target loss function to minimize energy consumption, maintain SOC balance, and improve driving smoothness, etc.

[0099] For example, the Adam algorithm can adjust the learning rate of the training parameters in the initial energy management model according to the historical information of the first moment (mean) and second moment (variance) of the gradient, so as to adapt to the sparsity or variability of different training parameters. The formula for its parameter update can be but is not limited to the following:

[0100]

[0101] where, g t is the gradient, and are the estimates of the momentum and variance of the gradient at time t, θ t+1 is the updated parameter, η is the learning rate, ∈ is introduced to avoid division by zero or close to zero, β 1 and β 2 are hyperparameters that control the decay rates of the first moment and second moment.

[0102] In the embodiments of the present application, an optimal power source torque distribution strategy and an engine start-stop control strategy can be directly generated from various data inputs of the vehicle by constructing a unified deep learning model and output. Thus, the use of end-to-end optimization eliminates the complex modular design and information transfer process in the traditional method, reduces the errors and computational burden in the intermediate links, and thus greatly improves the efficiency of energy management.

[0103] Optionally, in an embodiment of the present application, before end-to-end training the initial energy management model using a training data set, supervised learning labels, a target loss function, and a target optimizer, it further includes: determining an energy management optimization target corresponding to a target range-extended hybrid vehicle based on the energy management requirements corresponding to multiple range-extended hybrid vehicles; obtaining an optimization result according to the energy management optimization target, and determining the target loss function according to the optimization result.

[0104] In some embodiments, in order to achieve efficient energy management of range-extended hybrid vehicles, the present application can also determine its energy management optimization target based on the energy management requirements corresponding to multiple range-extended hybrid vehicles, thereby obtaining an optimization result according to the energy management optimization target, and further determining the target loss function.

[0105] Among them, the energy management requirement can be understood as the requirements of users or vehicle designers for vehicle energy management. For example, it is required to ensure the minimum energy consumption of the vehicle, or to give priority to ensuring the driving experience of users, etc. The energy management optimization target is to optimize the energy management of the vehicle while meeting these energy management requirements.

[0106] For example, in the present application, if the energy management requirements of multiple range-extended hybrid vehicles are to reduce energy consumption while ensuring the driving experience of users, the optimization target can be, but is not limited to, set as two parts: minimizing the total vehicle energy consumption and optimizing driving smoothness.

[0107] Among them, minimizing the total vehicle energy consumption can be considered to be achieved by optimizing fuel consumption and electric energy consumption, that is, by considering the power output of the engine and the motor, as well as information such as vehicle state, road conditions, and driving requirements, and dynamically adjusting the use of the power source, so as to achieve the minimum total vehicle energy consumption, which can be expressed in formula form as follows, but is not limited to:

[0108] J = ∫(C fuel + λ·C battery )dt

[0109] Among them, C fuel represents fuel consumption, and C battery represents electric energy consumption.

[0110] Moreover, the driving smoothness optimization in the embodiments of the present application can be achieved by reducing the frequent switching between the engine and the motor, that is, by smooth torque distribution to reduce the vibration and load change of the vehicle, which can be expressed in formula form but is not limited to the following:

[0111] min∑(|T engine,t+1 -T engine,t |+|T motor,t+1 -T motor,t |)

[0112] It should be noted that if the energy management requirement is only the minimization of the total vehicle energy consumption or the driving smoothness is given priority, different energy management optimization objectives can also be determined according to the corresponding energy management requirements.

[0113] After determining the energy management optimization objective, the embodiments of the present application can construct the objective loss function of the energy management model.

[0114] The embodiments of the present application can determine the energy management optimization objective according to the energy management requirement of the vehicle, and then determine the objective loss function of the model, thereby converting the objective of energy management into a quantifiable and optimizable mathematical form, providing a clear optimization direction for model training, and enabling the model to achieve the optimization of energy management by minimizing the loss function and improving performance indicators such as energy utilization efficiency.

[0115] Optionally, in an embodiment of the present application, before end-to-end training of the initial energy management model using the training data set, the supervised learning label, the objective loss function, and the objective optimizer, it further includes: constructing the energy consumption loss function, the power battery balance loss function, and the smoothness loss function of the objective loss function; determining the objective loss function based on the energy consumption loss function, the power battery balance loss function, and the smoothness loss function.

[0116] Based on the related descriptions of other embodiments, it can be understood that the embodiments of the present application can construct the objective loss function according to the energy management optimization objective.

[0117] In the actual execution process, when training the initial energy management model, in order to ensure that the energy management optimization objective can be achieved, the objective loss function L total constructed by the present application can be composed of, but is not limited to, the energy consumption loss function, the power battery balance loss function, and the smoothness loss function, and its formula can be expressed as, but is not limited to:

[0118] L total =L energy +L SOC-balance +L smoothness

[0119] wherein, L energy represents the energy consumption loss, and L SOC-balanceRepresents the SOC balance loss (power battery balance loss), L smoothness Represents the smoothness loss.

[0120] Among them, the construction expressions and meanings of each loss term can be expressed as follows:

[0121] (1) Energy consumption loss L energy is constructed to encourage the model to minimize the total energy consumption of the range-extended hybrid vehicle during driving, including but not limited to fuel consumption and battery power consumption. Its expression can be but not limited to the following:

[0122] L energy = λ fuel · C fuel + λ battery · C battery

[0123] Among them, C fuel represents fuel consumption, C battery represents power consumption, λ fuel 、λ battery are weighting coefficients used to balance the contributions of fuel consumption and power consumption in the total energy consumption.

[0124] (2) SOC balance loss L SOC is constructed to ensure that the battery operates within a reasonable range, avoid over-discharge or over-charge, and avoid frequent start-stop of the range extender resulting in low efficiency. Its expression can be but not limited to the following:

[0125] L SOC = α SoC · (ΔSOC) 2

[0126] In fact, α SOC is the weight coefficient, and ΔSOC represents the SOC deviation. Its formula can be but not limited to:

[0127] ΔSOC = SOC current - SOC target

[0128] Among them, SOC current represents the current battery SOC, and SOC target represents the target SOC.

[0129] In order to further constrain the change of SOC and avoid violent fluctuations in the battery SOC, the embodiment of the present application can also add a penalty term to constrain the change amount of SOC at two consecutive moments. Its formula can be but not limited to:

[0130] L SOC_change = β SOC · (SOCcurrent,t+1 -SOC current,t ) 2

[0131] wherein, β SOC is the weight coefficient of the SOC change penalty term.

[0132] Then the total SOC balance loss can be but is not limited to expressed as:

[0133] L SoC_balance = L SOC + L SOC_change

[0134] (3) The construction of the smoothness loss is to reduce the switching frequency of the power source, avoid switching the working states between the engine and the motor too frequently, so as to improve the driving smoothness, and its expression can be but is not limited to expressed as follows:

[0135]

[0136] wherein, γ engine and γ motor are weight coefficients, ΔT engine and ΔT motor respectively represent the engine torque change amount and the motor torque change amount, and its formula can be but is not limited to expressed as:

[0137] ΔT engine = |T engine,t+1 - T engine,t |

[0138] ΔT motor = |T motor,t+1 - T motor,t |

[0139] The embodiments of the present application can generate a target loss function by comprehensively considering various factors, can take into account different dimensions of goals at the same time, such as ensuring the minimum vehicle energy consumption while ensuring driving smoothness, and helps different loss functions to constrain the target energy management model from different angles, reduce the risk of overfitting, make the target energy management model perform more stably on complex and changeable data, and enhance the robustness and generalization ability of the target energy management model.

[0140] Optionally, in an embodiment of the present application, it further includes: optimizing a preset initial energy management model by using a learning rate decay strategy and an early stopping mechanism to construct an initial energy management model.

[0141] Based on the relevant descriptions of other embodiments, it can be understood that the embodiments of the present application can use a target optimizer such as the Adam optimizer for model training. Further, considering that in deep learning, the target energy management model may have the risk of overfitting, that is, it performs very well on the training data and can well fit various features and rules in the training data, but performs very poorly in new, unseen test data or actual application scenarios. The embodiments of the present application can also introduce a learning rate decay strategy and an early stopping mechanism to optimize the preset initial energy management model and avoid overfitting of the initial energy management model. Herein, the preset initial energy management model can be understood as the initial energy management model without introducing the learning rate decay strategy and the early stopping mechanism.

[0142] Specifically, as still shown in Figure 4 , in order to avoid jumping out of the local optimal solution, the embodiments of the present application can set a learning rate decay strategy in the energy management model. As the iteration progresses, the learning rate is gradually decreased, and its expression is as follows:

[0143]

[0144] where η t is the learning rate at the t-th iteration, η 0 is the initial learning rate, γ is the decay rate, and step is the decay period.

[0145] Next, the embodiments of the present application can introduce an early stopping mechanism by monitoring the validation set loss, setting patience parameters and thresholds. When the loss on the validation set no longer decreases or begins to increase, training is stopped in advance to prevent the initial energy management model from overfitting the training data.

[0146] And, Table 1 is the training parameter setting table of an embodiment of the present application and can be as follows:

[0147] Table 1

[0148]

[0149]

[0150] The embodiments of the present application can introduce a learning rate decay strategy and an early stopping mechanism to optimize the preset initial energy management model and avoid overfitting of the initial energy management model, so as to ensure that the obtained target energy management model after training is adaptable to new, unseen test data, thereby ensuring the performance of the target energy management model in actual application scenarios.

[0151] The energy management method for a range-extended hybrid vehicle applied in the model construction stage according to an embodiment of the present application can collect various current vehicle information, various information of the future target driving section, and the predicted average speed, and directly generate the torque distribution result and the engine start-stop control result through the trained target energy management model, and then generate the corresponding control instructions to control the target range-extended hybrid vehicle to achieve energy management. Thus, the initial energy management model is trained and optimized end-to-end using a large amount of data sets, supervised learning labels, target loss functions, and target optimizers to obtain the target energy management model, enabling the target energy management model in the present application to directly input information and output the optimal power source torque distribution and engine start-stop control, eliminating the complex modular design and information transmission process in traditional methods, reducing the errors and computational burdens in intermediate links, thereby greatly improving the energy management efficiency, contributing to achieving full-process optimization. At the same time, while ensuring that the obtained power source torque distribution result is a globally optimal energy management strategy, the target energy management model in the present application has the reasoning ability to perform real-time energy management on the target range-extended hybrid vehicle. When facing complex and changeable road conditions and environmental conditions, it can quickly respond and adjust the energy management strategy, with extremely high practicality while ensuring the robustness and generalization of the target energy management model. Thus, it solves the problems in related technologies that rule-based methods are difficult to adapt to dynamic environments, the control accuracy of model prediction is limited by model quality, both have certain limitations under complex and changeable road conditions, and the separate implementation of the energy consumption prediction model and the energy management control strategy easily leads to low information transmission efficiency and difficulty in integrating full-process optimization, etc.

[0152] Next, a description is given with reference to the drawings of an energy management device for a range-extended hybrid vehicle applied in the model construction stage according to an embodiment of the present application.

[0153] Figure 5 It is a schematic structural diagram of an energy management device for a range-extended hybrid vehicle according to an embodiment of the present application.

[0154] As Figure 5 shown, the energy management device 10 of the range-extended hybrid vehicle includes: a collection module 100, an extraction module 200, a generation module 300, and a first construction module 400.

[0155] Among them, the collection module 100 is used to collect the operation information, environmental information, and actual energy output information of multiple range-extended hybrid vehicles at at least one historical moment, as well as the road condition information and predicted average speed of the target driving section of multiple range-extended hybrid vehicles after at least one historical moment;

[0156] An extraction module 200 is configured to extract static features corresponding to multiple range-extended electric vehicles at at least one historical moment according to operation information and environmental information, extract temporal features corresponding to a target driving section after at least one historical moment of multiple range-extended electric vehicles according to road condition information and predicted average vehicle speed, and construct a training data set according to the static features, temporal features, and actual energy output information;

[0157] A generation module 300 is configured to generate motor torque distribution results, engine torque distribution results, and engine start-stop control results of multiple range-extended electric vehicles at at least one historical moment based on the training data set, so as to generate supervised learning labels of an initial energy management model;

[0158] A first construction module 400 is configured to end-to-end train an initial energy management model by using the training data set, supervised learning labels, a target loss function, and a target optimizer, so as to construct a target energy management model for managing the energy of a target range-extended electric vehicle.

[0159] Optionally, in an embodiment of the present application, it further includes: a first determination module and a second determination module.

[0160] The first determination module is configured to determine an energy management optimization target of the target range-extended electric vehicle based on the energy management requirements of the target range-extended electric vehicle before end-to-end training the initial energy management model by using the training data set, supervised learning labels, a target loss function, and a target optimizer;

[0161] The second determination module is configured to obtain an optimization result according to the energy management optimization target, and determine a target loss function according to the optimization result.

[0162] Optionally, in an embodiment of the present application, it further includes: a second construction module and a third determination module.

[0163] The second construction module is configured to construct an energy consumption loss function, a power battery balance loss function, and a smoothness loss function of the target loss function before end-to-end training the initial energy management model by using the training data set, supervised learning labels, a target loss function, and a target optimizer;

[0164] The third determination module is configured to determine a target loss function based on the energy consumption loss function, the power battery balance loss function, and the smoothness loss function.

[0165] Optionally, in an embodiment of the present application, it further includes: a third construction module, configured to optimize a preset initial energy management model by using a learning rate decay strategy and an early stopping mechanism, so as to construct an initial energy management model.

[0166] It should be noted that the foregoing explanation of the embodiment of the energy management method for a range-extended hybrid vehicle applied to the model construction stage is also applicable to the energy management device for a range-extended hybrid vehicle in which this embodiment is applied to the model construction stage, and details are not described herein again.

[0167] According to the energy management device for a range-extended hybrid vehicle applied to the model construction stage provided by an embodiment of the present application, by collecting various current vehicle information, various information of a future target driving section, and a predicted average speed, a torque distribution result and an engine start-stop control result can be directly generated through a trained target energy management model, and then corresponding control instructions are generated to control the target range-extended hybrid vehicle to achieve energy management. Thus, an initial energy management model is trained and optimized end-to-end by using a large amount of data sets, supervised learning labels, a target loss function, and a target optimizer to obtain a target energy management model, so that the target energy management model in the present application can directly input information to output an optimal power source torque distribution and engine start-stop control, eliminating the complex modular design and information transmission process in the traditional method, reducing the errors and computational burdens in the intermediate links, thereby greatly improving the energy management efficiency, contributing to the realization of the whole-process optimization. At the same time, while ensuring that the obtained power source torque distribution result is a globally optimal energy management strategy, the inference ability of the target energy management model in the present application can achieve real-time energy management of the target range-extended hybrid vehicle. When facing complex and changeable road conditions and environmental conditions, it can quickly respond and adjust the energy management strategy, with extremely high practicality while ensuring the robustness and generalization of the target energy management model. Thus, the problems in the related art are solved, that is, the rule-based method is difficult to adapt to the dynamic environment, the control accuracy based on model prediction is limited by the model quality, both have certain limitations under complex and changeable road conditions, and the separate implementation of the energy consumption prediction model and the energy management control strategy is likely to lead to low information transmission efficiency and difficulty in integrating the whole-process optimization.

[0168] The above embodiments describe the model construction stage. The embodiments of the model application stage will be described below.

[0169] Specifically, Figure 6 is a flowchart of an energy management method for a range-extended hybrid vehicle applied to the model application stage provided by an embodiment of the present application.

[0170] As Figure 6 shown, the energy management method for the range-extended hybrid vehicle is applied to the model construction stage and includes the following steps:

[0171] In step S601, real-time operation information and real-time environment information of the target range-extended hybrid vehicle are collected, and road condition information and predicted average speed of the target range-extended hybrid vehicle on a future target driving section are collected.

[0172] During the actual execution process, the embodiments of the present application can collect the input information required by the target energy management model, that is, the real-time operation information and real-time environment information of the target range-extended hybrid vehicle, and collect the road condition information and predicted average speed of the target range-extended hybrid vehicle on the future target driving section. Herein, the target range-extended hybrid vehicle can be understood as a range-extended hybrid vehicle that applies the energy management method of the range-extended hybrid vehicle in the present application to achieve vehicle energy management.

[0173] For example, the present application can collect the real-time throttle pedal opening, real-time brake pedal opening, real-time battery SOC, real-time engine speed, real-time motor speed, real-time vehicle speed, real-time environmental temperature of the target range-extended hybrid vehicle; and, the predicted average speed of the target range-extended hybrid vehicle on the future target driving section and information such as the slope, length, and congestion degree of the future target driving section.

[0174] Herein, the future target driving section refers to a driving section of a certain duration or a certain length in the future based on the current moment. It should be noted that the selection rule of the future target driving section and the calculation rule of the predicted average speed are the same as the selection rule of the target driving section and the calculation rule of the predicted average speed mentioned above.

[0175] In step S602, the real-time operation information, real-time environment information, road condition information, and predicted average speed are input into the pre-constructed target energy management model, the real-time static characteristics of the target range-extended hybrid vehicle and the temporal characteristics of the target range-extended hybrid vehicle on the future target driving section are extracted, and the static characteristics and temporal characteristics are fused to obtain a fused feature, so as to output the real-time motor torque distribution result, real-time engine torque distribution result, and real-time engine start-stop control result of the target range-extended hybrid vehicle according to the fused feature.

[0176] In some embodiments, after collecting the information required by the target energy management model, the embodiments of the present application can input the real-time operation information, real-time environment information, road condition information, and predicted average speed into the pre-constructed target energy management model.

[0177] The feature extraction module in the target energy management model can use this information to extract the static characteristics of the target range-extended hybrid vehicle at the current moment and the temporal characteristics on the future target driving section, and then use the feature fusion module in the target energy management model to fuse the static characteristics and temporal characteristics to obtain a fused feature, so as to generate the real-time motor torque distribution result, real-time engine torque distribution result, and real-time engine start-stop control result of the target range-extended hybrid vehicle at the current moment according to the fused feature, and output the result using the output module in the target energy management model.

[0178] In step S603, corresponding control instructions are generated according to the real-time motor torque distribution result, the real-time engine torque distribution result, and the real-time engine start-stop control result to control the target range-extended hybrid vehicle for energy management.

[0179] In some embodiments, Figure 7 is a flowchart of the actual application of a model according to an embodiment of the present application. As Figure 7 shown:

[0180] Step S701, collect information on the vehicle and the future target driving section in real time;

[0181] Step S702, input the collected information into the trained target energy management model for torque distribution, and output the real-time motor torque distribution result, the real-time engine torque distribution result, and the real-time engine start-stop control result;

[0182] Step S703, generate corresponding control instructions according to these results;

[0183] Step S704, send these control instructions to the corresponding components of the target range-extended hybrid vehicle, and the components execute the control instructions to finally realize the energy management of the target range-extended hybrid vehicle.

[0184] The energy management method for a range-extended hybrid vehicle applied in the model application stage according to the embodiments of the present application can collect various current vehicle information, various information of the future target driving section, and the predicted average speed, and directly generate the torque distribution result and the engine start-stop control result through the trained target energy management model, and then generate the corresponding control instructions to control the target range-extended hybrid vehicle to achieve energy management. Thus, it realizes end-to-end training and optimization of the initial energy management model using a large amount of data sets, supervised learning labels, target loss functions, and target optimizers to obtain the target energy management model, enabling the target energy management model in the present application to directly input information and output the optimal power source torque distribution and engine start-stop control, eliminating the complex modular design and information transmission process in the traditional method, reducing the errors and computational burden in the intermediate links, thereby greatly improving the efficiency of energy management, contributing to the realization of the whole-process optimization. Moreover, while ensuring that the obtained power source torque distribution result is a globally optimal energy management strategy, the target energy management model in the present application has the inference ability to achieve real-time energy management of the target range-extended hybrid vehicle. When facing complex and changeable road conditions and environmental conditions, it can quickly respond and adjust the energy management strategy, with extremely high practicality while ensuring the robustness and generalization of the target energy management model. Thus, it solves the problems in the related technologies that the rule-based method is difficult to adapt to the dynamic environment, the control accuracy of the model prediction is limited by the model quality, both have certain limitations under complex and changeable road conditions, and the energy consumption prediction model and the energy management control strategy are separately implemented, which easily leads to low information transmission efficiency and difficulty in integrating the whole-process optimization, etc.

[0185] Next, refer to the drawings to describe the energy management device for a range-extended hybrid vehicle applied in the model application stage according to the embodiments of the present application.

[0186] Figure 8 It is a schematic structural diagram of the energy management device for a range-extended hybrid vehicle applied in the model application stage according to the embodiments of the present application.

[0187] As Figure 8 shown, the energy management device 20 for a range-extended hybrid vehicle applied in the model application stage includes: a collection module 500, a processing module 600, and a management module 700.

[0188] Among them, the collection module 500 is used to collect the real-time operation information and real-time environment information of the target range-extended hybrid vehicle, and collect the road condition information and predicted average speed of the target range-extended hybrid vehicle on the future target driving section.

[0189] A processing module 600 is configured to input real-time operation information, real-time environment information, road condition information, and predicted average vehicle speed into a pre-constructed target energy management model, extract real-time static features of the target range-extended hybrid vehicle and sequential features of the target range-extended hybrid vehicle on a future target driving section, and fuse the static features and sequential features to obtain fused features, so as to output real-time motor torque distribution results, real-time engine torque distribution results, and real-time engine start-stop control results of the target range-extended hybrid vehicle according to the fused features.

[0190] A management module 700 is configured to generate corresponding control instructions according to the real-time motor torque distribution results, real-time engine torque distribution results, and real-time engine start-stop control results, so as to control the target range-extended hybrid vehicle to perform energy management.

[0191] It should be noted that the foregoing explanation of the energy management method embodiment of the range-extended hybrid vehicle applied to the model application stage is also applicable to the energy management device of the range-extended hybrid vehicle applied to the model application stage of this embodiment, and will not be elaborated here.

[0192] According to the energy management device of the range-extended hybrid vehicle applied to the model application stage proposed in the embodiments of the present application, by collecting various information of the vehicle at present, various information of the future target driving section, and predicted average vehicle speed, torque distribution results and engine start-stop control results can be directly generated through a trained target energy management model, and then corresponding control instructions are generated to control the target range-extended hybrid vehicle to achieve energy management. Thus, it is realized to train and optimize an initial energy management model end-to-end by using a large amount of data sets, supervised learning labels, a target loss function, and a target optimizer to obtain a target energy management model, so that the target energy management model in the present application can directly input information and output the optimal power source torque distribution and engine start-stop control, eliminating the complex modular design and information transmission process in the traditional method, reducing the errors and calculation burden in the intermediate links, thereby greatly improving the energy management efficiency, contributing to the realization of the whole process optimization. At the same time, while ensuring that the obtained power source torque distribution result is a globally optimal energy management strategy, the target energy management model in the present application has an inference ability to realize real-time energy management of the target range-extended hybrid vehicle. When facing complex and changeable road conditions and environmental conditions, it can quickly respond and adjust the energy management strategy, with extremely high practicality while ensuring the robustness and generalization of the target energy management model. Thus, the problems in the related technologies are solved, that is, the rule-based method is difficult to adapt to the dynamic environment, the control accuracy based on model prediction is limited by the model quality, both have certain limitations under complex and changeable road conditions, and the energy consumption prediction model and the energy management control strategy are separately implemented, which easily leads to low information transmission efficiency and difficulty in integrating the whole process optimization.

[0193] Figure 9The structural schematic diagram of the vehicle provided by the embodiment of the present application. The vehicle may include:

[0194] A memory 901, a processor 902, and a computer program stored on the memory 901 and executable on the processor 902.

[0195] When the processor 902 executes the program, it implements the energy management method of the range-extended hybrid vehicle provided in the above embodiment.

[0196] Furthermore, the vehicle further includes:

[0197] A communication interface 903 for communication between the memory 901 and the processor 902.

[0198] The memory 901 is used to store a computer program executable on the processor 902.

[0199] The memory 901 may include a high-speed RAM memory, and may also include non-volatile memory, such as at least one disk memory.

[0200] If the memory 901, the processor 902, and the communication interface 903 are implemented independently, the communication interface 903, the memory 901, and the processor 902 may be interconnected through a bus and communicate with each other. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 9 only a thick line is shown in the figure, but it does not mean that there is only one bus or one type of bus.

[0201] Optionally, in a specific implementation, if the memory 901, the processor 902, and the communication interface 903 are integrated on a chip, the memory 901, the processor 902, and the communication interface 903 may communicate with each other through an internal interface.

[0202] The processor 902 may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0203] The embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the energy management method of the range-extended hybrid vehicle as described above is implemented.

[0204] The embodiments of the present application also provide a computer program product, including a computer program. The computer program can run computer instructions, and when the computer instructions are executed by a processor, the energy management method of the range-extended hybrid vehicle provided by the embodiments of the present application is implemented.

[0205] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0206] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "N" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0207] Any process or method description shown in the flowchart or described in other ways herein can be understood as representing a module, segment, or part of the code including one or N executable instructions for implementing a customized logic function or process. The scope of the preferred embodiments of the present application includes additional implementations, where the functions can be executed in a manner that is not shown or discussed, including in a substantially simultaneous manner according to the involved functions or in a reverse order, which should be understood by those skilled in the art of the embodiments of the present application.

[0208] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definable sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in conjunction with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion (electronic device) having one or N wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0209] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or combinations thereof. In the above-described embodiments, the N steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented using hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having suitable combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), and the like.

[0210] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the methods of the above-described embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0211] In addition, each functional unit in various embodiments of the present application may be integrated into a processing module, may exist physically alone for each unit, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0212] The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disc, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. An energy management method for a range-extended hybrid vehicle, characterized in that: Applied to the model building stage, wherein the method comprises the following steps: Collecting operation information, environmental information, and actual energy output information of a plurality of extended-range hybrid vehicles at at least one historical moment, as well as road condition information and predicted average vehicle speeds of the plurality of extended-range hybrid vehicles at a target driving section after the at least one historical moment; Extracting static features corresponding to the plurality of extended-range hybrid vehicles at the at least one historical moment according to the operation information and the environmental information, extracting temporal features corresponding to a target driving section of the plurality of extended-range hybrid vehicles after the at least one historical moment according to the road condition information and the predicted average vehicle speed, and constructing a training data set according to the static features, the temporal features and the actual energy output information; Based on the training data set, generating motor torque distribution results, engine torque distribution results, and engine start-stop control results of the plurality of extended-range hybrid vehicles at the at least one historical moment to generate supervised learning labels for an initial energy management model; The initial energy management model is trained end-to-end using the training data set, the supervised learning labels, the target loss function, and the target optimizer to construct a target energy management model for managing the energy of a target extended-range hybrid vehicle.

2. The method according to claim 1, characterized in that Before end-to-end training the initial energy management model using the training data set, the supervised learning label, the target loss function and the target optimizer, the method further includes: Determining energy management optimization targets corresponding to the plurality of extended-range hybrid vehicles based on energy management requirements corresponding to the plurality of extended-range hybrid vehicles; According to the energy management optimization target, an optimization result is obtained, and the target loss function is determined according to the optimization result.

3. The method according to claim 1, characterized in that Before end-to-end training the initial energy management model using the training data set, the supervised learning label, the target loss function and the target optimizer, the method further includes: Constructing an energy consumption loss function, a power battery balance loss function and a smoothness loss function of the target loss function; The target loss function is determined based on the energy consumption loss function, the power battery balancing loss function and the smoothness loss function.

4. The method according to claim 1, characterized in that: Also includes: The preset initial energy management model is optimized by using a learning rate decay strategy and an early stopping mechanism to construct the initial energy management model.

5. An energy management method for a range-extended hybrid vehicle, characterized in that: Applied in the model application stage, the energy management method of the extended-range hybrid vehicle according to any one of claims 1 to 4 is adopted, wherein the method comprises the following steps: Collecting real-time operating information and real-time environmental information of the target extended-range hybrid vehicle, and collecting road condition information and predicted average vehicle speed of the target extended-range hybrid vehicle in a future target driving section; Inputting the real-time operation information, the real-time environment information, the road condition information and the predicted average vehicle speed into a pre-constructed target energy management model, extracting the real-time static features of the target extended-range hybrid vehicle and the time series features of the target extended-range hybrid vehicle in a future target driving section, and fusing the static features and the time series features to obtain fused features, so as to output the real-time motor torque distribution result, the real-time engine torque distribution result and the real-time engine start-stop control result of the target extended-range hybrid vehicle according to the fused features; Corresponding control instructions are generated according to the real-time motor torque distribution result, the real-time engine torque distribution result and the real-time engine start-stop control result to control the target extended-range hybrid vehicle to perform energy management.

6. An energy management device for a range-extended hybrid vehicle, characterized in that: Applied to the model building stage, wherein the device comprises: a collection module, configured to collect operation information, environmental information, and actual energy output information of a plurality of extended-range hybrid vehicles at at least one historical moment, as well as road condition information and predicted average vehicle speeds of the plurality of extended-range hybrid vehicles at a target driving section after the at least one historical moment; an extraction module, configured to extract static features corresponding to the plurality of extended-range hybrid vehicles at the at least one historical moment according to the operation information and the environmental information, extract temporal features corresponding to a target driving section of the plurality of extended-range hybrid vehicles after the at least one historical moment according to the road condition information and the predicted average vehicle speed, and construct a training data set according to the static features, the temporal features and the actual energy output information; A generating module, configured to generate, based on the training data set, motor torque distribution results, engine torque distribution results, and engine start-stop control results of the plurality of extended-range hybrid vehicles at the at least one historical moment, so as to generate a supervised learning label for an initial energy management model; A construction module is used to train the initial energy management model end-to-end using the training data set, the supervised learning label, the target loss function and the target optimizer to construct a target energy management model for managing the energy of a target extended-range hybrid vehicle.

7. An energy management device for a range-extended hybrid vehicle, characterized in that: Applied in the model application stage, the energy management device of the extended-range hybrid vehicle as claimed in claim 5 is used, wherein the device comprises: A collection module, used to collect real-time operation information and real-time environmental information of the target extended-range hybrid vehicle, and to collect road condition information and predicted average vehicle speed of the target extended-range hybrid vehicle in a future target driving section; a processing module, configured to input the real-time operation information, the real-time environment information, the road condition information and the predicted average vehicle speed into a pre-constructed target energy management model, extract the real-time static features of the target extended-range hybrid vehicle and the time series features of the target extended-range hybrid vehicle in a future target driving section, and fuse the static features and the time series features to obtain fused features, so as to output a real-time motor torque distribution result, a real-time engine torque distribution result and a real-time engine start-stop control result of the target extended-range hybrid vehicle according to the fused features; The management module is used to generate corresponding control instructions according to the real-time motor torque distribution result, the real-time engine torque distribution result and the real-time engine start-stop control result, so as to control the target extended-range hybrid vehicle to perform energy management.

8. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the energy management method for a range-extended hybrid vehicle as described in any one of claims 1 to 4 or 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the energy management method for a range-extended hybrid vehicle as described in any one of claims 1-4 or 5.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed, it is used to implement the energy management method for the extended-range hybrid vehicle as described in any one of claims 1-4 or 5.

Images