A cloud-based method for updating the economical driving torque of a distributed electric vehicle

Abstract

This invention provides a cloud-based method for updating the economical driving torque of a distributed drive electric vehicle. The method includes: analyzing driver intent under preset simplified conditions to obtain a generalized demand driving torque; defining torque distribution coefficients to allocate the generalized demand driving torque to the front and rear motors of the electric vehicle; calculating the total driving efficiency of the drive system based on the efficiency characteristics of each motor under driving conditions; obtaining the torque distribution coefficient that maximizes the total driving efficiency of the electric vehicle at the current accelerator pedal opening and current motor speed, and calculating an optimal torque distribution map based on the torque distribution coefficients; performing preliminary calculations on the optimal torque distribution map to obtain a data stream, and then using the cloud to sequentially clean and train the data stream to complete the cloud-based update process for the electric vehicle. This invention can process streaming data and perform real-time optimization, which helps to optimize the torque distribution of distributed drive electric vehicles, thereby increasing driving economy.

Description

Technical Field

[0001] This invention relates to the field of automotive energy management technology, and in particular to a cloud-based method for updating the economical driving torque of a distributed drive electric vehicle. Background Technology

[0002] In recent years, in response to the new energy crisis and the ecological environment crisis, the new energy vehicle industry has attracted much attention due to its high resource efficiency. Compared with traditional centralized drive electric vehicles, front and rear axle independent drive electric vehicles have significant advantages in vehicle handling and maneuverability. Through reasonable torque configuration, fuel economy and power can be greatly improved. How to more rationally distribute the economic torque of pure electric four-wheel drive electric vehicles is a key research direction for universities and automakers.

[0003] For the economical distribution of driving torque in pure electric four-wheel drive vehicles, the mainstream approach is to target the highest drive system efficiency, use the motor efficiency diagram for offline numerical optimization, and employ an traversal optimization algorithm to obtain a set of optimal torque distribution coefficient curves for the front and rear axles, so that the motor operates at the best efficiency point. In the real-time control system, the optimal distribution coefficient is obtained by looking up a table based on the current total torque demand and motor speed.

[0004] However, traditional allocation schemes generate offline data diagrams, which are unsuitable for operating conditions with varying parameters. The relationship between speed, torque, and optimal motor efficiency is highly sensitive to many environmental factors, such as temperature and tire wear, which can cause the optimal torque allocation coefficients to deviate. Therefore, the optimized allocation coefficients obtained based on offline diagrams cannot always guarantee the optimal efficiency of the drive system. Summary of the Invention

[0005] Based on this, the purpose of the present invention is to provide a cloud-based method for updating the economical driving torque of a distributed drive electric vehicle, so as to at least solve the shortcomings of the above-mentioned related technologies.

[0006] This invention proposes a cloud-based method for updating the economical driving torque of a distributed drive electric vehicle, comprising:

[0007] The driver's intention is analyzed under preset simplified conditions to obtain the driver's generalized demand driving torque;

[0008] Define torque distribution coefficients to distribute the generalized demand drive torque to the front and rear motors of the electric vehicle;

[0009] The total driving efficiency of the electric vehicle's drive system is calculated based on the efficiency characteristics of each motor of the electric vehicle under preset driving conditions.

[0010] Obtain the torque distribution coefficient that maximizes the overall driving efficiency of the drive system when the electric vehicle is at the current accelerator pedal opening and the current motor speed, and calculate the optimal torque distribution diagram of the drive system based on the torque distribution coefficient;

[0011] The optimal torque distribution map of the drive system is initially calculated to obtain the corresponding data stream. The data stream is then cleaned and trained sequentially using the cloud to complete the cloud update process of the economical drive torque of the electric vehicle.

[0012] Furthermore, the step of analyzing the driver's intention under preset simplified conditions to obtain the driver's generalized demand driving torque includes:

[0013] The driver's intention is analyzed using the linear accelerator pedal analysis method to obtain the linear relationship between the torque load coefficient of the electric motor and the opening of the accelerator pedal in the electric vehicle.

[0014] The torque load coefficient of the electric motor is calculated based on the linear relationship, and the generalized driving torque required by the current driver is calculated using the torque load coefficient and the external characteristics of the front and rear axle motors in the electric motor.

[0015] Furthermore, the expression for the linear relationship between the torque load coefficient of the electric motor and the accelerator pedal opening is as follows:

[0016] L D =100%·Acc_Pedal;

[0017] In the formula, L D The torque load coefficient of the motor under driving conditions; Acc_Pedal is the accelerator pedal opening.

[0018] The formula for calculating the generalized demand driving torque of the current driver is as follows:

[0019]

[0020] T req =T req，1 ·η1+T req，2 ·η2;

[0021] In the formula, i = 1 and 2 represent the front and rear motors, respectively, and T req，i For the required torque of each motor, T req Driven by the general needs of drivers, T max，i n represents the peak torque corresponding to different motors. i For different motor speeds, n b，i η represents the base speed corresponding to different motors, and η is the reduction ratio.

[0022] Furthermore, the expression for the torque distribution coefficient is as follows:

[0023]

[0024] In the formula, T f and T r These represent the torques of the front and rear motors, respectively.

[0025] Furthermore, the step of calculating the total driving efficiency of the electric vehicle's drive system based on the efficiency characteristics of each motor under preset driving conditions includes:

[0026] The efficiency characteristics of each motor in the electric vehicle are calculated based on the system efficiency of each motor and the efficiency of the reducer.

[0027] The torque output of the front and rear motors is calculated based on the efficiency characteristics of each motor in the electric vehicle and the efficiency of the transmission system of the front and rear motors.

[0028] The total input power and total output power of the electric vehicle's drive system are calculated based on the torque output of the front and rear motors, and the total drive efficiency of the drive system is obtained based on the total input power and the total output power.

[0029] Furthermore, the formula for calculating the torque output of the front and rear motors is as follows:

[0030]

[0031] η mf =η mr =η r ;

[0032]

[0033] In the formula, n mf n mr These are the rotational speeds of the front and rear motors, η. mf η mr The efficiency of the drive system for the front and rear motors, T, are respectively. mf T mr These represent the output torques of the front and rear motors, respectively.

[0034] Furthermore, the formula for calculating the total driving efficiency of the electric vehicle's drive system is as follows:

[0035]

[0036]

[0037]

[0038] In the formula, Pm_in P m_out Let be the total input power and total output power of the drive system of the electric vehicle, respectively, and η be the total drive efficiency of the drive system, where:

[0039]

[0040] In the formula, T max_f T max_r At rotational speed n m The maximum torque of the front and rear motors, n mf_max n mr_max These represent the maximum speeds corresponding to the front and rear motors, P. bat_max η is the maximum charge / discharge power of the battery. dis This refers to the battery's charging and discharging efficiency.

[0041] Compared with existing technologies, the beneficial effects of this invention are as follows: It utilizes online graph learning to perform real-time optimal control of the distribution of driving torque under various operating conditions, generating an online-updated 3D graph, which helps to respond more quickly to transient operating conditions; it employs a permutation model and stochastic gradient algorithm to iteratively update the learned map, adapting to changing operating conditions and updating the graph under transient or steady-state operating conditions, reducing computational complexity and memory usage; it adopts a cloud-based update strategy, transferring the optimization process to the cloud, reducing onboard computation and memory requirements, and directly obtaining the updated graph; it can process streaming data and perform real-time optimization, which helps to optimize the torque distribution of distributed drive electric vehicles, thereby increasing driving economy. Attached Figure Description

[0042] Figure 1 This is a flowchart of the cloud-based update method for the economical driving torque of a distributed-drive electric vehicle in the first embodiment of the present invention;

[0043] Figure 2 for Figure 1 Detailed flowchart of step S101;

[0044] Figure 3 for Figure 1 Detailed flowchart of step S103;

[0045] Figure 4 This is a three-dimensional diagram of the drive torque distribution coefficient in the first embodiment of the present invention;

[0046] Figure 5 This is a pixel map of the driving torque distribution coefficient in the first embodiment of the present invention;

[0047] Figure 6 This is a diagram of the OTA system architecture in the first embodiment of the present invention.

[0048] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0049] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0050] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0051] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0052] Example 1

[0053] Please see Figure 1 The figure shows a cloud-based method for updating the economic driving torque of a distributed drive electric vehicle according to the first embodiment of the present invention. The method specifically includes steps S101 to S105:

[0054] S101, Analyze the driver's intention under preset simplified conditions to obtain the driver's generalized demand driving torque;

[0055] For further details, please refer to Figure 2 Step S101 specifically includes steps S1011 to S1012:

[0056] S1011, The driver's intention is analyzed using the linear accelerator pedal analysis method to obtain the linear relationship between the torque load coefficient of the electric motor and the opening of the accelerator pedal of the electric vehicle.

[0057] S1012, calculate the torque load coefficient of the electric motor based on the linear relationship, and calculate the generalized driving torque required by the current driver using the torque load coefficient and the external characteristics of the front and rear axle motors in the electric motor.

[0058] In practical implementation, to simplify the model and calculations, the following assumptions are made first:

[0059] (1) When a car is traveling in a straight line on a level road with good road conditions, the road gradient is considered to be 0.

[0060] (2) The four wheels rotate at the same speed at all times;

[0061] (3) The driving torque is first distributed to the front and rear axles in a certain proportion, and then evenly distributed to the left and right wheels.

[0062] Specifically, to simplify the simulation process, the linear accelerator pedal analysis method is used, which states that under driving conditions, there is a certain linear relationship between the torque load coefficient of the electric motor and the opening of the accelerator pedal:

[0063] L D =100%·Acc_Pedal;

[0064] In the formula, L D The torque load coefficient of the motor under driving conditions; Acc_Pedal is the accelerator pedal opening.

[0065] Starting from the accelerator pedal opening, the motor torque load coefficient is calculated, and combined with the external characteristics of the front and rear axle motors, the generalized demand driving torque of the current driver is obtained.

[0066]

[0067] T req =T req，1 ·η1+T req，2 ·η2;

[0068] In the formula, i = 1 and 2 represent the front and rear motors, respectively, and T req，i For the required torque of each motor, T req Driven by the general needs of drivers, T max，i n represents the peak torque corresponding to different motors. i For different motor speeds, n b，i η represents the base speed corresponding to different motors, and η is the reduction ratio.

[0069] S102, Define torque distribution coefficients to distribute the generalized demand drive torque to the front and rear motors of the electric vehicle;

[0070] In practical implementation, a torque distribution coefficient K is defined, which is the ratio of the torque allocated to the front wheel hub to the total required torque, distributing the generalized required torque to the front and rear axle motors. Generally speaking, the value of the torque distribution coefficient is in the range of [0, 1], where 0 represents rear-wheel drive and 1 represents front-wheel drive.

[0071]

[0072] In the formula, Tf and T r Let represent the torques of the front and rear motors, respectively; where the torques of the front and rear motors satisfy the following formula:

[0073] T req =T f +T r .

[0074] S103, calculate the total driving efficiency of the electric vehicle's drive system based on the efficiency characteristics of each motor of the electric vehicle under preset driving conditions.

[0075] For further details, please refer to Figure 3 Step S103 specifically includes steps S1031 to S1033:

[0076] S1031, calculate the efficiency characteristics of each motor of the electric vehicle based on the system efficiency basis of each motor and the efficiency of the reducer.

[0077] S1032, calculate the torque output of the front and rear motors based on the efficiency characteristics of each motor of the electric vehicle and the efficiency of the transmission system of the front and rear motors.

[0078] S1033, calculate the total input power and total output power of the drive system of the electric vehicle based on the torque output of the front and rear motors, and obtain the total drive efficiency of the drive system based on the total input power and the total output power.

[0079] In practical implementation, the total drive efficiency of the drive system can be calculated based on the efficiency characteristics of each motor under the provided drive conditions. Since the torque transmission of the centralized motor passes through a reducer, the system efficiency value of the centralized motor is multiplied by the reducer efficiency to obtain the efficiency characteristic of the centralized motor system.

[0080] From the following formula:

[0081]

[0082] It can be seen that when the vehicle speed is the same, the front and rear motors rotate at the same speed.

[0083] Set n mf =n mr =n m The efficiency η of the front and rear axle drive system mf =η mr =η r Then the torque output of the front and rear motors is:

[0084]

[0085] In the formula, n mf nmr These are the rotational speeds of the front and rear motors, η. mf η mr The efficiency of the drive system for the front and rear motors, T, are respectively. mf T mr These represent the output torques of the front and rear motors, respectively.

[0086] Therefore, it can be seen that there is a certain functional relationship between motor efficiency and motor speed and torque:

[0087]

[0088] At the current speed, due to the different torque distribution coefficients, the operating points of the front and rear axle motors differ, resulting in variations in the efficiency of each motor and consequently, the efficiency of the entire transmission system. Therefore, by adjusting the torque distribution coefficients, the operating states of the front and rear axle motors can be adjusted to maximize the driving efficiency of the entire system, thereby achieving the optimal distribution of total braking force.

[0089] Therefore, the total input power (kW) of the dual-motor four-wheel drive system, i.e., the total power received from the battery, is:

[0090]

[0091] The total output power when driven by two motors is:

[0092]

[0093] By combining the formulas for total input power and total output power, the overall efficiency of the drive system under driving conditions can be obtained:

[0094]

[0095] In the formula, P m_in P m_out Let be the total input power and total output power of the drive system of the electric vehicle, respectively, and η be the total drive efficiency of the drive system, where:

[0096]

[0097] In the formula, T max_f T max_r At rotational speed n m The maximum torque of the front and rear motors, n mf_max n mr_max These represent the maximum speeds corresponding to the front and rear motors, P. bat_max η is the maximum charge / discharge power of the battery. dis This refers to the battery's charging and discharging efficiency.

[0098] S104, obtain the torque distribution coefficient that maximizes the total driving efficiency of the drive system when the electric vehicle is at the current accelerator pedal opening and the current motor speed, and calculate the optimal torque distribution diagram of the drive system based on the torque distribution coefficient;

[0099] Based on the above, Matlab was used to program the calculations, deriving the torque distribution coefficient that maximizes the overall efficiency of the drive system under the current accelerator pedal opening (representing the driver's demand for driving force) and motor speed. This coefficient was then used as a lookup table model for control strategy modeling, thereby reducing the model's complexity. The calculation results based on a front-to-rear axle distributed drive electric vehicle model are as follows: Figures 4 to 5 As shown, under different motor speeds and different accelerator pedal openings, the torque distribution coefficient is also different when the overall efficiency of the vehicle's transmission system is maximized.

[0100] (1) In the range of low vehicle speed and small accelerator pedal opening, the torque distribution coefficient is 1. The results show that when driving at low speed and with low torque demand, using a single front axle drive can achieve a higher overall drive system efficiency;

[0101] (2) As the vehicle speed gradually increases to medium-high speeds, the torque distribution coefficient is almost zero when the accelerator pedal opening is small. The results show that under high-speed driving conditions, using the rear axle drive alone can achieve a higher overall drive system efficiency;

[0102] (3) Within the range of larger accelerator pedal opening, the torque distribution coefficient gradually transitions from 1 to 0.5, indicating that when the required torque is large, using a combined front and rear axle drive can achieve higher transmission efficiency. The torque distribution coefficient surface exhibits some fluctuations, which is determined by the motor efficiency characteristics and the constrained range of the torque distribution coefficient.

[0103] (4) When both the accelerator pedal opening and the engine speed are at intermediate values, which is also the case in most situations, the torque distribution coefficient is basically fixed at around 0.5. The results show that in most cases, when the vehicle speed and pedal opening are moderate, using a combined front and rear axle drive can achieve higher transmission efficiency.

[0104] S105, the optimal torque distribution diagram of the drive system is initially calculated to obtain the corresponding data stream, and the data stream is cleaned and trained sequentially using the cloud to complete the cloud update process of the economical drive torque of the electric vehicle.

[0105] In practical implementation, to adjust the torque distribution coefficient to ensure optimal economy, we obtained the optimal torque distribution coefficient diagram for a specific speed and torque in the previous step. However, the data in this diagram requires calibration and can be offset due to engine aging, changes in operating conditions, etc. The traditional method for obtaining the grid values ​​in this diagram is calibration, even if the motor runs for multiple cycles under stable conditions on a test bench; however, the motor may operate under transient conditions, and the actual operating point is not always located at the grid points in the diagram.

[0106] The parameters of this 3D diagram are: torque distribution coefficient K, total drive efficiency n, and speed n. m and the output torque T of the front axle motor mf The rotational speed represents the motor's operating condition; the rotational speed n in the diagram... m and the output torque T of the front axle motor mf It is discrete, thus establishing the mapping {n} m T mf}→{K,η}, each grid point records two values, namely the optimal torque distribution coefficient and the corresponding total drive efficiency. Based on this three-dimensional graph, the optimal torque distribution coefficient can be obtained by interpolation, thus obtaining {K,η} corresponding to each operating point.

[0107] In the actual operation of the motor, the operating point may not be strictly located on the standard grid point. The value of such a point can be calculated by linear interpolation, that is, by using the values ​​of adjacent grid points to perform multivariate interpolation on the grid points of the conventional three-dimensional graph.

[0108] For the sake of versatility and convenience, the three dimensions of the 3D graph are written as x1, x2, x3. Then (i, j, l) are the normalized coordinate values, l is the value K or η contained in the grid point, and the interpolation point (i+u, j+v, l+q), 0≤u, v, q<1 can be linearly approximated in the local space.

[0109]

[0110] To calibrate the calculated 3D graph, which may change grid values ​​due to factors such as motor aging and changes in operating conditions, the traditional calibration method is to run the motor through multiple working cycles under stable operating conditions. However, in reality, the motor may operate under unstable transient conditions, and the actual operating point may not be within the 3D graph. Therefore, a self-learning method is used to update the graph to solve the above problems.

[0111] A graph-based cloud-based online update strategy is proposed, enabling graph updates under any operating condition, not just the stable conditions of a test bench, thus eliminating a significant amount of tedious calibration work. Furthermore, the cloud-based update method stores streaming data in the cloud and performs data cleaning and filtering, eliminating the need to store large amounts of data in the vehicle's ECU.

[0112] Using the training set {x} obtained from the cloud n o n}, n = 1, 2, ..., N, o n For the observed values, η=φ T W is the model that needs to be learned from the training data. The training goal is to find the model parameters W that minimize the loss function J.

[0113]

[0114] J n (W) is the sample {x} n o n Learning objectives;

[0115]

[0116] Where, φ n =φ(x n Derive the partial derivatives of the model parameter W:

[0117]

[0118] e n Let {x} represent the deviation between the observed values ​​and the model output. n o n If n = 1, 2, ..., N is the training data stream, then the samples will be used one by one for training, and the parameters W will be updated using stochastic gradient descent.

[0119] W n+1 =W n +γφ n e n ;

[0120] γ is the step size of stochastic gradient descent, and the convergence of online learning is guaranteed by 0 < γ < 2. Mapping {n m T mf}→K and {n m T mf}→η can be learned using the algorithms described above.

[0121] Specifically, to achieve cloud-based system updates, an OTA (Over-the-air) system is built, with the following architecture: Figure 6As shown, the entire system mainly consists of an OTA cloud server, OTA terminals, and OTA objects. The OTA cloud server's structure primarily comprises five parts: an OTA management platform, OTA upgrade service, task scheduling, file service, and task management. It supports multiple security encryption and decryption algorithms and includes complete upgrade packages for all controllers supporting OTA upgrades. The OTA terminal mainly includes an OTA engine and an OTA adapter. The OTA engine acts as a bridge connecting the OTA terminal and the OTA cloud, enabling secure communication between the cloud and the terminal, including upgrade package download, upgrade package decryption, and differential package reconstruction. The OTA adapter is a different implementation encapsulated according to unified interface requirements to ensure compatibility with different update logics or processes of different software or devices. The upgrade adapter is provided by the respective ECU software implementations that require OTA upgrades. OTA objects are the on-vehicle controllers that support OTA, mainly including cockpit controllers, ADAS controllers, and in-vehicle embedded controllers. In this invention, it primarily involves the drive torque distribution controller.

[0122] First, vehicle model and information are entered into the OTA cloud information management system. An initial optimal torque distribution map is calculated and transmitted to the vehicle. The OTA object collects new information such as four-wheel torque and energy consumption, calculates the corresponding front-to-rear axle torque distribution ratio and efficiency over a period of time, and uploads the data stream to the OTA cloud server. The cloud server cleans the data stream and uses the preprocessed data stream for training, i.e., {x} n o n}, n = 1, 2, ..., N. The trained and updated torque distribution map is periodically updated to the vehicle, thus completing the cloud-based training and updating process for economical driving torque.

[0123] In summary, the cloud-based update method for the economical driving torque of distributed drive electric vehicles in the above embodiments of the present invention utilizes online graph learning to perform real-time optimal control of the distribution of driving torque under various operating conditions, generating an online-updated 3D graph, which helps to respond more quickly to transient operating conditions; by employing a linear interpolation model and stochastic gradient algorithm to iteratively update the learned map, it can adapt to changing operating conditions, updating the graph under transient or steady-state operating conditions, reducing computational complexity and memory usage; by adopting a cloud-based update strategy, the optimization process is transferred to the cloud, reducing on-board computation and memory requirements, and directly obtaining the updated graph; it can process streaming data and perform real-time optimization, which helps to optimize the torque distribution of distributed drive electric vehicles, thereby increasing driving economy.

[0124] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0125] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A cloud-based method for updating the economical driving torque of a distributed-drive electric vehicle, characterized in that, include: The driver's intention is analyzed under preset simplified conditions to obtain the driver's generalized demand driving torque; Define torque distribution coefficients to distribute the generalized demand drive torque to the front and rear motors of the electric vehicle; The total driving efficiency of the electric vehicle's drive system is calculated based on the efficiency characteristics of each motor of the electric vehicle under preset driving conditions. Obtain the torque distribution coefficient that maximizes the overall driving efficiency of the drive system when the electric vehicle is at the current accelerator pedal opening and the current motor speed, and calculate the optimal torque distribution diagram of the drive system based on the torque distribution coefficient; The optimal torque distribution map of the drive system is initially calculated to obtain the corresponding data stream. The data stream is then cleaned and trained sequentially using cloud computing to complete the cloud update process for the economical drive torque of the electric vehicle. The parameters of the optimal torque distribution map are: torque distribution coefficient K, total drive efficiency η, and rotational speed n. m and the output torque T of the front axle motor mf The rotational speed n m The operating condition of the motor, the speed n m and the output torque T of the front axle motor mf For discrete data, the steps of performing preliminary calculations on the optimal torque distribution map of the drive system to obtain the corresponding data stream, and then using the cloud to sequentially perform data cleaning and data training on the data stream to complete the cloud update process of the economical drive torque of the electric vehicle include: Establish mapping {n m T mf }→{K, η}, each grid point records two values: the optimal torque distribution coefficient and the corresponding total drive efficiency. Based on the optimal torque distribution map, the optimal torque distribution coefficient is obtained by interpolation, thus obtaining {K, η} corresponding to each operating point. If the optimal torque distribution map is written as x1, x2, x3, then (i, j, l) are the normalized coordinate values, l is the value K or η contained in the grid point, and the interpolation point (i+u, j+v, l+q), 0≤u, v, q<1, can be linearly approximated in the local space. Using the training set {x} obtained from the cloud n o n }, n = 1, 2, ..., N, o n For the observed values, η=φ T • W is the model that needs to be learned from the training data. The training objective is to find the model parameters W that minimize the loss function J. J n (W) is the sample {x} n ,o n Learning objectives; Where, φ n =φ(x n Derive the partial derivatives of the model parameter W: e n Let {x} represent the deviation between the observed values ​​and the model output. n o n If n = 1, 2, ..., N is the training data stream, then the samples will be used for training one by one, and the parameters W will be updated using stochastic gradient descent. IN n+1 =In n +γφ n e n ; γ is the step size of stochastic gradient descent. The convergence of online learning is guaranteed by 0 < γ < 2, mapping {n m ,T mf }→K and {n m ,T mf }→η can be learned using the algorithms described above.

2. The cloud-based update method for the economical driving torque of a distributed drive electric vehicle according to claim 1, characterized in that, The steps for analyzing the driver's intentions under preset simplified conditions to obtain the driver's generalized demand driving torque include: The driver's intention is analyzed using the linear accelerator pedal analysis method to obtain the linear relationship between the torque load coefficient of the electric motor and the opening of the accelerator pedal in the electric vehicle. The torque load coefficient of the electric motor is calculated based on the linear relationship, and the generalized driving torque required by the current driver is calculated using the torque load coefficient and the external characteristics of the front and rear axle motors in the electric motor.

3. The cloud-based update method for the economical driving torque of a distributed drive electric vehicle according to claim 2, characterized in that, The expression for the linear relationship between the torque load coefficient of the electric motor and the accelerator pedal opening is as follows: L D =100%·Acc_Pedal; In the formula, L D The torque load coefficient of the motor under driving conditions; Acc_Pedal is the accelerator pedal opening. The formula for calculating the generalized demand driving torque of the current driver is as follows: T req =T req，1 ·η1+T req，2 ·η2; In the formula, i = 1, 2 represent the front and rear motors respectively, and T req，i For the required torque of each motor, T req Driven by the general needs of drivers, T max，i n represents the peak torque corresponding to different motors. i For different motor speeds, n b，i η represents the base speed corresponding to different motors, and η is the reduction ratio.

4. The cloud-based update method for the economical driving torque of a distributed drive electric vehicle according to claim 3, characterized in that, The expression for the torque distribution coefficient is: In the formula, T f and T r These represent the torques of the front and rear motors, respectively.

5. The cloud-based update method for the economical driving torque of a distributed drive electric vehicle according to claim 4, characterized in that, The steps for calculating the total driving efficiency of the electric vehicle's drive system based on the efficiency characteristics of each motor under preset driving conditions include: The efficiency characteristics of each motor in the electric vehicle are calculated based on the system efficiency of each motor and the efficiency of the reducer. The torque output of the front and rear motors is calculated based on the efficiency characteristics of each motor in the electric vehicle and the efficiency of the transmission system of the front and rear motors. The total input power and total output power of the electric vehicle's drive system are calculated based on the torque output of the front and rear motors, and the total drive efficiency of the drive system is obtained based on the total input power and the total output power.

6. The cloud-based update method for the economical driving torque of a distributed drive electric vehicle according to claim 5, characterized in that, The formula for calculating the torque output of the front and rear motors is as follows: or mf =the mr =the r ; In the formula, n mf η mr These represent the rotational speeds of the front and rear motors, η. mf η mr The efficiency of the drive system for the front and rear motors, respectively, T mf T mr These represent the output torque of the front and rear motors, respectively.

7. The cloud-based update method for the economical driving torque of a distributed drive electric vehicle according to claim 6, characterized in that, The formula for calculating the total driving efficiency of the electric vehicle's drive system is as follows: In the formula, P m_in P m_out Let be the total input power and total output power of the drive system of the electric vehicle, respectively, and η be the total drive efficiency of the drive system, where: In the formula, T max_f T max_r At rotational speed n m The maximum torque of the front and rear motors, n mf_max n mr_max These represent the maximum speeds of the front and rear motors, respectively. bat_max η is the maximum charge / discharge power of the battery. dis This refers to the battery's charge and discharge efficiency.

Images