A vehicle adaptive coasting energy feedback control method and device and vehicle

Abstract

The application relates to the technical field of new energy vehicles, and provides a vehicle adaptive coasting energy feedback control method and device and a vehicle. The method comprises the following steps: collecting real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, real-time road slope and current coasting recovery torque of a driving motor when the vehicle is in a coasting energy feedback control mode; inputting the real-time vehicle speed, the real-time vehicle weight, the real-time road friction coefficient and the real-time road slope into a preset target recovery torque prediction model to output a target coasting recovery torque; and if the absolute value of a torque difference value between the current coasting recovery torque and the target coasting recovery torque is greater than or equal to a preset torque threshold value, the current coasting recovery torque of the driving motor is adaptively adjusted so that the absolute value of the torque difference value is less than the preset torque threshold value. The target coasting recovery torque obtained by the application is relatively accurate, and the energy recovery of the vehicle under a coasting working condition can be maximized, thereby reducing energy waste.

Description

Technical Field

[0001] This application relates to the field of new energy vehicle technology, and in particular to a vehicle adaptive coasting energy feedback control method, device and vehicle. Background Technology

[0002] New energy vehicles generally have two modes of energy recovery: braking energy recovery and coasting energy recovery. Braking energy recovery is achieved by pressing the brake pedal, while coasting energy recovery is achieved by coasting in neutral.

[0003] Current coasting energy recovery control strategies all rely on a lookup table method based on vehicle speed to obtain the recovered torque. However, this method fails to consider other factors affecting recovered torque besides vehicle speed (such as vehicle weight and road friction coefficient), leading to inaccurate recovered torque readings. Inaccurate recovered torque can cause drivers to misjudge deceleration, affecting driving safety, and also prevents the maximum utilization of vehicle inertia for energy recovery, resulting in energy waste. Summary of the Invention

[0004] In view of this, embodiments of this application provide a vehicle adaptive coasting energy feedback control method, device and vehicle to solve the problems in the prior art where inaccurate recovery torque can easily lead to driver misjudgment of deceleration, thus affecting driving safety, and at the same time, it cannot maximize the utilization of vehicle inertia to recover energy, resulting in energy waste.

[0005] A first aspect of this application provides a vehicle adaptive coasting energy feedback control method, comprising:

[0006] When the vehicle is in coasting energy recovery control mode, the real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, real-time road gradient, and current coasting recovery torque of the drive motor are collected.

[0007] Input the real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient and real-time road slope into the preset target recovery torque prediction model, and output the target coasting recovery torque.

[0008] Calculate the torque difference between the current coasting recovery torque and the target coasting recovery torque;

[0009] If the absolute value of the torque difference is greater than or equal to the preset torque threshold, the current coasting recovery torque of the drive motor will be adaptively adjusted so that the absolute value of the torque difference is less than the preset torque threshold.

[0010] A second aspect of this application provides a vehicle adaptive coasting energy feedback control device, comprising:

[0011] The data acquisition module is configured to collect the vehicle's real-time speed, real-time vehicle weight, real-time road friction coefficient, real-time road gradient, and current coasting recovery torque when the vehicle is in coasting energy recovery control mode.

[0012] The output module is configured to input real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient and real-time road slope into a preset target recovery torque prediction model, and output the target coasting recovery torque.

[0013] The calculation module is configured to calculate the torque difference between the current coasting recovery torque and the target coasting recovery torque;

[0014] The adjustment module is configured to adaptively adjust the current coasting recovery torque if the absolute value of the torque difference is greater than or equal to a preset torque threshold, so that the absolute value of the torque difference is less than the preset torque threshold.

[0015] A third aspect of the embodiments of this application provides a vehicle, including a coasting energy recovery control system, the coasting energy recovery control system including a recovery controller and a drive motor;

[0016] The feedback controller includes the vehicle adaptive coasting energy feedback control device described in the second aspect above.

[0017] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method.

[0018] Compared with the prior art, the beneficial effects of this application embodiment include at least the following: by comprehensively considering various factors affecting the coasting energy feedback control when the vehicle is in coasting energy feedback control mode (such as vehicle speed, vehicle weight, road friction coefficient and road slope), and using a preset target recovery torque prediction model to predict, the target coasting recovery torque is obtained. The target coasting recovery torque obtained in this way is more accurate and can help the driver correctly judge the deceleration of the vehicle, thereby improving driving safety. Then, by calculating the torque difference between the target coasting recovery torque and the current coasting recovery torque of the drive motor, the current coasting recovery torque of the drive motor is adaptively adjusted according to the torque difference, so that the absolute value of the torque difference is less than the preset torque threshold. This can achieve a smooth transition of power limitation during vehicle shifting while maximizing coasting energy recovery when the vehicle is in coasting energy feedback control mode, thereby reducing energy waste. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a block diagram of a gliding energy feedback control system according to an embodiment of this application;

[0021] Figure 2 This is a schematic flowchart of a vehicle adaptive coasting energy feedback control method provided in an embodiment of this application;

[0022] Figure 3 This is a schematic diagram of the structure of a vehicle adaptive coasting energy feedback control device provided in an embodiment of this application;

[0023] Figure 4 This is a schematic diagram of the structure of a vehicle provided in an embodiment of this application;

[0024] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0025] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0026] The following describes in detail, with reference to the accompanying drawings, a vehicle adaptive coasting energy feedback control method, device, and vehicle according to embodiments of this application.

[0027] Figure 1 This is a block diagram of a gliding energy feedback control system according to an embodiment of this application. Figure 1 As shown, the gliding energy feedback control system includes a feedback controller 101 and a drive motor 102.

[0028] The feedback controller 101 can be a controller (e.g., a microcontroller) that is set separately outside the drive motor 102, or it can be a drive motor controller set inside the drive motor, or it can be a remote controller, etc.

[0029] The feedback controller 101 can determine whether the vehicle is in a coasting condition by acquiring the accelerator pedal opening signal and the brake pedal opening signal. If both the accelerator pedal opening signal and the brake pedal opening signal are fully open, the vehicle is determined to be in a coasting condition, and the vehicle enters the coasting energy feedback control mode. Coasting condition refers to the condition in which the vehicle gradually decelerates by relying on the reverse generated torque of the drive motor and various resistances when both the accelerator pedal opening and the brake pedal opening are fully open.

[0030] When the vehicle is in coasting energy recovery control mode, the recovery controller 101 collects the vehicle's real-time speed, real-time vehicle weight, real-time road friction coefficient, real-time road slope, and the current coasting recovery torque of the drive motor 102; inputs the real-time speed, real-time vehicle weight, real-time road friction coefficient, and real-time road slope into a preset target recovery torque prediction model, and outputs the target coasting recovery torque; calculates the torque difference between the current coasting recovery torque and the target coasting recovery torque; if the absolute value of the torque difference is greater than or equal to a preset torque threshold, the current coasting recovery torque of the drive motor 102 is adaptively adjusted so that the absolute value of the torque difference is less than the preset torque threshold.

[0031] The technical solution provided in this application comprehensively considers various factors affecting coasting energy recovery control when the vehicle is in coasting energy recovery control mode (such as vehicle speed, vehicle weight, road friction coefficient, and road slope), and uses a preset target recovery torque prediction model to predict and obtain the target coasting recovery torque. The target coasting recovery torque obtained in this way is relatively accurate and can help the driver correctly judge the vehicle's deceleration, thereby improving driving safety. Then, by calculating the torque difference between the target coasting recovery torque and the current coasting recovery torque of the drive motor, the current coasting recovery torque of the drive motor is adaptively adjusted according to the torque difference, so that the absolute value of the torque difference is less than a preset torque threshold. This can achieve a smooth transition of power limitation during vehicle shifting while maximizing coasting energy recovery in the coasting energy recovery control mode, thereby reducing energy waste.

[0032] In another embodiment, the coasting energy recovery control system may include a feedback controller, a motor controller, and a drive motor. The feedback controller, when the vehicle is in coasting energy recovery control mode, collects the vehicle's real-time speed, real-time vehicle weight, real-time road friction coefficient, real-time road gradient, and the current coasting recovery torque of the drive motor; inputs the real-time speed, real-time vehicle weight, real-time road friction coefficient, and real-time road gradient into a preset target recovery torque prediction model, outputs the target coasting recovery torque, and sends the target coasting recovery torque to the motor controller. The motor controller receives the target coasting recovery torque and calculates the torque difference between the target coasting recovery torque and the current coasting recovery torque; when the absolute value of the torque difference is greater than or equal to a preset torque threshold, it adaptively adjusts the current coasting recovery torque of the drive motor so that the absolute value of the torque difference is less than the preset torque threshold.

[0033] Figure 2 This is a flowchart illustrating a vehicle adaptive coasting energy feedback control method provided in an embodiment of this application. Figure 2 The vehicle adaptive coasting energy feedback control method can be derived from Figure 1 The feedback controller 101 executes. For example... Figure 2 As shown, the adaptive coasting energy feedback control method for this vehicle includes:

[0034] Step S201: When the vehicle is in coasting energy recovery control mode, collect the vehicle's real-time speed, real-time vehicle weight, real-time road friction coefficient, real-time road gradient, and the current coasting recovery torque of the drive motor.

[0035] The feedback controller 101 can obtain the current coasting recovery torque of the drive motor by reading the torque parameters of the drive motor.

[0036] Step S202: Input the real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient and real-time road slope into the preset target recovery torque prediction model, and output the target coasting recovery torque.

[0037] In some embodiments, the target recovery torque prediction model is trained by the following steps:

[0038] Collect a sample dataset, which includes multiple sample data. Each sample data includes historical vehicle parameters and maximum historical coasting recovery torque when the vehicle or the vehicle and its similar vehicles are in coasting energy recovery control mode. The historical vehicle parameters include historical vehicle speed, historical vehicle weight, and historical road gradient.

[0039] The sample dataset is divided into a training dataset and a test dataset. The training dataset includes multiple training data points, and the test dataset includes multiple test data points. Each training data point and test data point is a single sample data point from the multiple sample data points.

[0040] The first training model is obtained by training multiple training data from the training dataset to train the preset initial recovery torque prediction model.

[0041] Input at least one test data point from the test dataset into the first training model and output the first predicted value of the recovery torque;

[0042] The first loss function value is calculated based on the first predicted recovery torque value and the actual recovery torque value of the test data input into the first training model;

[0043] If the value of the first loss function meets the set error range, then the first training model is determined as the target recovery torque prediction model.

[0044] Generally, car models include small cars, microcars, compact cars, mid-range cars, high-end cars, luxury cars, sedans, CDVs, MPVs, SUVs, and other types of vehicles.

[0045] Vehicles of the same type refer to vehicles belonging to the same category as the vehicle currently requiring adaptive coasting energy regeneration control. For example, if the vehicle currently requiring adaptive coasting energy regeneration control is a small car, then vehicles of the same type are also small cars.

[0046] Due to differences in vehicle weight and structure, the coasting energy recovery control modes of different vehicle models may vary significantly. In a preferred embodiment of this application, a target recovery torque prediction model can be trained for vehicles of the same model (e.g., a target recovery torque prediction model M1 is trained for small cars, a target recovery torque prediction model M2 is trained for micro cars, and so on). In practical use, the corresponding target recovery torque prediction model can be selected based on the vehicle model to which adaptive coasting energy recovery control is performed, and the corresponding target coasting recovery torque can be output based on the target recovery torque prediction model, thereby improving the accuracy of the obtained target coasting recovery torque.

[0047] As an example, assuming the sample dataset includes 1000 sample data points, 800 of these 1000 sample data points can be allocated as the training dataset, and the remaining 200 as the test dataset. Then, these 800 training data points are used to train a pre-defined initial recovery torque prediction model, resulting in the first training model.

[0048] The initial recovery torque prediction model can be either a convolutional neural network (CNN) model or a deep neural network (DNN) model. A CNN model includes an input layer, convolutional layers, fully connected layers, and an output layer. A DNN model includes an input layer, an output layer, multiple hidden layers, and a loss function engine, with the hidden layers arranged along the forward propagation path between the input and output layers.

[0049] In one embodiment, when the initial recovery torque prediction model is a convolutional neural network model, the training process of the preset initial recovery torque prediction model is performed using multiple training data from the training dataset to obtain the first training model is as follows:

[0050] Multiple training data points from the training dataset are input into the input layer of the convolutional neural network model. The input data is then passed to the convolutional layer, where it undergoes convolution processing to obtain convolutional features. These features are then output to a fully connected layer, and subsequently to the output layer, where the output is obtained. Next, a cross-entropy loss function is calculated based on this output. The weights of each neuron in the fully connected layer are adjusted using this cross-entropy loss function until a preset convergence condition is met, at which point training stops, resulting in the first trained model.

[0051] In another embodiment, when the initial recovery torque prediction model is a deep neural network model, the training process of the preset initial recovery torque prediction model is performed using multiple training data from the training dataset to obtain the first training model is as follows:

[0052] First, multiple training data points from the training dataset are input into the input layer of the deep neural network model. Then, these multiple training data points are output to at least one hidden layer among multiple hidden layers via the forward propagation path. At least some of these hidden layers include quaternion layers, which are used to perform consistent quaternion forward operations (including historical vehicle speed, historical vehicle weight, historical road friction coefficient, and historical road slope) based on one or more variable parameters to generate at least one corresponding feature map along the forward propagation path and output it to the output layer. The output layer is used to generate a DNN (Deep Neural Network) result based on ternary forward operations and outputs the DNN result to the loss function engine. The loss function engine calculates a loss function (such as root mean square error) based on the DNN result and uses the loss function for backpropagation to train the initial recovery torque prediction model to obtain the first training model.

[0053] At least one test data point is randomly selected from the test dataset and input into the first training model trained by the above steps, outputting a first predicted recovery torque value. Assuming the test data input to the first training model is test data 01, the output first predicted recovery torque value is x1, and the actual recovery torque value corresponding to test data 01 is x2, then based on x1, x2, and a preset first loss function (e.g., cross-entropy loss function), the first loss function value loss1 is calculated. If the first loss function loss1 satisfies a preset error range (e.g., ≤σ), then the first training model is determined as the target recovery torque prediction model.

[0054] In some embodiments, if the first loss function value does not meet the set error range, the first model parameters of the first training model are updated and adjusted using the first loss function value to obtain the first updated model.

[0055] The test data that was not input into the first training model is used as the new training data, or a new training dataset is collected and the first updated model is trained using the new training data to obtain the second training model.

[0056] If the accuracy of the second training model meets the preset accuracy range, then the second training model will be determined as the target recovery torque prediction model.

[0057] For ease of understanding, continuing with the example above, assume that the first loss function value loss1 > σ, meaning it does not meet the given error range. Then, the first model parameters (including the weights and biases of the first trained model) are updated and adjusted using the first loss function value loss1, resulting in the first updated model. Next, the first updated model is trained using test data not input to the first trained model (i.e., test data other than test data 01), or using training data from a newly collected training dataset, to obtain the second trained model. The training process of the first updated model is similar to that of the first trained model described above, and will not be repeated here.

[0058] The preset accuracy range can be flexibly set according to the actual situation. For example, it can be set to ≥0.8, ≥0.9, etc.

[0059] If the preset accuracy range is ≥0.8, and the accuracy of the second training model is 0.95 (≥0.8, which meets the preset accuracy range), then the second training model will be determined as the target recovery torque prediction model.

[0060] In other embodiments, if the first loss function value does not meet the set error range, the first model parameters of the first training model are updated and adjusted using the first loss function value to obtain the first updated model.

[0061] The first updated model is trained using multiple training data points from the training dataset to obtain the third training model;

[0062] Input at least one test data point from the test dataset into the third training model and output the second predicted value of the recovery torque.

[0063] The second loss function value is calculated based on the second predicted recovery torque value and the actual recovery torque value of the test data input into the third training model;

[0064] If the value of the second loss function meets the set error range, then the third training model will be determined as the target recovery torque prediction model.

[0065] To facilitate understanding, let's continue with the example above. Assume the first loss function value, loss1, is greater than σ, meaning it doesn't meet the given error range. Then, we use the first loss function value, loss1, to update and adjust the first model parameters (including the weights and biases) of the first trained model, resulting in the first updated model. Next, we continue training the first updated model using multiple training data points from the training dataset, resulting in the third trained model.

[0066] The second loss function value, loss2, can be calculated using the same method as the first loss function value described above, and will not be repeated here. If the second loss function value, loss2, meets the set error range (e.g., ≤σ), then the third training model is determined as the target recovery torque prediction model. If the second loss function value, loss2, does not meet the set error range, then the model parameters of the third training model are updated and adjusted using the second loss function value, loss2, to obtain the second updated model. Then, multiple training data from the training dataset are used to train the second updated model to obtain the fourth training model, and the third loss function value, loss3, is calculated. If the third loss function value, loss3, does not meet the set error range, then the above training process is repeated until the loss function value meets the set error range, or the number of training rounds meets the preset round threshold (e.g., 50 rounds, 100 rounds, etc.), then the training ends, and the model obtained in the last round of training (or the training model with the best model accuracy) is determined as the target recovery torque prediction model.

[0067] In some embodiments, step S202 above specifically includes:

[0068] Based on real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, and real-time road gradient, calculate the vehicle's rolling resistance, air resistance, slope resistance, and acceleration resistance.

[0069] The target coasting recovery torque of the vehicle is determined based on rolling resistance, air resistance, slope resistance, and acceleration resistance.

[0070] Specifically, the rolling resistance of a vehicle can be calculated based on its real-time vehicle weight and real-time road friction coefficient; the air resistance of a vehicle can be calculated based on its real-time vehicle speed, air resistance coefficient, projected area of ​​the vehicle in the direction of travel, and air density; the slope resistance of a vehicle can be calculated based on its real-time road gradient and real-time vehicle weight; and the acceleration resistance of a vehicle can be calculated based on its real-time vehicle speed and real-time vehicle weight.

[0071] The rolling resistance F of the vehicle is calculated according to formula (1). f .

[0072] F f =W*f (1)

[0073] In equation (1), W represents the sum of wheel loads (i.e., real-time vehicle weight), and f represents the real-time road friction coefficient.

[0074] Calculate the vehicle's air resistance F according to formula (2). w .

[0075] Fw = 0.5 * CD * A * ρ * u r 2 (2)

[0076] In equation (2), CD represents the air resistance coefficient, A represents the frontal area, i.e., the projected area in the direction of vehicle travel, ρ represents the air density, and u r This indicates relative speed, which is the speed of a car in calm weather (i.e., real-time vehicle speed).

[0077] Calculate the vehicle's gradient resistance F according to formula (3). i .

[0078] F i =Gsinα (3)

[0079] In equation (3), G represents the gravity acting on the car, G = mg, m represents the weight of the car, g represents the gravitational acceleration, and α represents the real-time road slope.

[0080] Ignoring the influence of the rotational inertia of the motor, the acceleration resistance F of the vehicle can be calculated according to formula (4). j .

[0081] F j =ma (4)

[0082] In equation (4), a represents the vehicle's acceleration, which is related to the real-time vehicle speed, and m represents the vehicle's weight.

[0083] The vehicle's coasting recovery torque F is calculated according to formula (5). m .

[0084] Fm =F f +F w +F i +F j (5).

[0085] Based on the rolling resistance F of the vehicle f air resistance F w , ramp resistance F i Acceleration resistance F j Historical recovery torque is used to establish a torque mapping relationship, that is, various resistances (involving vehicle weight m, road friction coefficient f, vehicle speed v, and gradient α) and the vehicle's coasting recovery torque F. m The mapping relationship between them, F m =f(m,f,v,α). The target coasting recovery torque refers to the coasting recovery torque when the vehicle's recovery power is at its maximum, i.e., F. m =max f(m,f,v,α).

[0086] Step S203: Calculate the torque difference between the current coasting recovery torque and the target coasting recovery torque.

[0087] In one embodiment, the torque difference can be calculated according to the following formula: Torque difference = Current coasting recovery torque - Target coasting recovery torque.

[0088] Step S204: If the absolute value of the torque difference is greater than or equal to a preset torque threshold, the current coasting recovery torque of the drive motor is adaptively adjusted so that the absolute value of the torque difference is less than the preset torque threshold. Specifically, the current coasting recovery torque can be decreased or increased within a preset time interval according to a preset torque gradient, so that the absolute value of the torque difference between the current coasting recovery torque and the target coasting recovery torque is less than the preset torque threshold.

[0089] The preset torque threshold can be flexibly set according to actual conditions, for example, it can be set to 100 Nm, 150 Nm, etc. It is usually set to 100 Nm.

[0090] The preset time interval can be flexibly set according to actual needs, such as 400 milliseconds, 500 milliseconds, etc. It is usually set to 500 milliseconds.

[0091] The preset torque gradient reflects the rate of torque change, that is, the amount of torque change per unit time.

[0092] As an example, assuming the preset torque threshold is 100 Nm and the preset time interval is 500 milliseconds, the absolute value of the torque difference is 200 Nm. This torque difference (200 Nm) is greater than the preset torque threshold (100 Nm). In this case, the current coasting recovery torque of the drive motor can be adaptively adjusted within 500 milliseconds according to the preset torque gradient, so that the absolute value of the torque difference is less than 100 Nm.

[0093] The technical solution provided in this application comprehensively considers various factors (such as vehicle speed, vehicle weight, road friction coefficient, and road gradient) that affect the coasting energy feedback control mode when the vehicle is in coasting energy feedback control mode. It then uses a preset target recovery torque prediction model to predict and obtain the target coasting recovery torque. The obtained target coasting recovery torque is relatively accurate and can help the driver correctly judge the vehicle's deceleration, thereby improving driving safety. Subsequently, by calculating the torque difference between the target coasting recovery torque and the current coasting recovery torque of the drive motor, the current coasting recovery torque of the drive motor is adaptively adjusted based on the torque difference, so that the absolute value of the torque difference is less than a preset torque threshold. This allows for a smooth transition of power limitation during vehicle shifting while maximizing coasting energy recovery in the coasting energy feedback control mode.

[0094] All of the above-mentioned optional technical solutions can be combined in any way to form the optional embodiments of this application, and will not be described in detail here.

[0095] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0096] Figure 3 This is a schematic diagram of a vehicle adaptive coasting energy feedback control device provided in an embodiment of this application. Figure 3 As shown, the vehicle's adaptive coasting energy recovery control device includes:

[0097] The data acquisition module 301 is configured to acquire the vehicle's real-time speed, real-time vehicle weight, real-time road friction coefficient, real-time road gradient, and current coasting recovery torque when the vehicle is in coasting energy recovery control mode.

[0098] Output module 302 is configured to input real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient and real-time road slope into a preset target recovery torque prediction model, and output the target coasting recovery torque.

[0099] The calculation module 303 is configured to calculate the torque difference between the current coasting recovery torque and the target coasting recovery torque;

[0100] The adjustment module 304 is configured to adaptively adjust the current coasting recovery torque if the absolute value of the torque difference is greater than or equal to a preset torque threshold, so that the absolute value of the torque difference is less than the preset torque threshold.

[0101] The technical solution provided in this application comprehensively considers various factors affecting coasting energy recovery control when the vehicle is in coasting energy recovery control mode (such as vehicle speed, vehicle weight, road friction coefficient, and road slope), and uses a preset target recovery torque prediction model to predict and obtain the target coasting recovery torque. The target coasting recovery torque obtained in this way is relatively accurate and can help the driver correctly judge the vehicle's deceleration, thereby improving driving safety. Then, by calculating the torque difference between the target coasting recovery torque and the current coasting recovery torque of the drive motor, the current coasting recovery torque of the drive motor is adaptively adjusted according to the torque difference, so that the absolute value of the torque difference is less than a preset torque threshold. This can achieve a smooth transition of power limitation during vehicle shifting while maximizing coasting energy recovery in the coasting energy recovery control mode, thereby reducing energy waste.

[0102] In some embodiments, the target recovery torque prediction model is trained by the following steps:

[0103] Collect a sample dataset, which includes multiple sample data. Each sample data includes historical vehicle parameters and maximum historical coasting recovery torque when the vehicle or a vehicle and its similar vehicles are in coasting energy recovery control mode. The historical vehicle parameters include historical vehicle speed, historical vehicle weight, historical road friction coefficient, and historical road gradient.

[0104] The sample dataset is divided into a training dataset and a test dataset. The training dataset includes multiple training data points, and the test dataset includes multiple test data points. Each training data point and test data point is a single sample data point from the multiple sample data points.

[0105] The first training model is obtained by training multiple training data from the training dataset to train the preset initial recovery torque prediction model.

[0106] Input at least one test data point from the test dataset into the first training model and output the first predicted value of the recovery torque;

[0107] The first loss function value is calculated based on the first predicted recovery torque value and the actual recovery torque value of the test data input into the first training model;

[0108] If the value of the first loss function meets the set error range, then the first training model is determined as the target recovery torque prediction model.

[0109] In some embodiments, after calculating the first loss function value based on the first predicted recovery torque value and the actual recovery torque value of the test data input to the first training model, the method further includes:

[0110] If the first loss function value does not meet the set error range, the first model parameters of the first training model are updated and adjusted using the first loss function value to obtain the first updated model.

[0111] The test data that was not input into the first training model is used as the new training data, or a new training dataset is collected and the first updated model is trained using the new training data to obtain the second training model.

[0112] If the accuracy of the second training model meets the preset accuracy range, then the second training model will be determined as the target recovery torque prediction model.

[0113] In some embodiments, after calculating the first loss function value based on the first predicted recovery torque value and the actual recovery torque value of the test data input to the first training model, the method further includes:

[0114] If the first loss function value does not meet the set error range, the first model parameters of the first training model are updated and adjusted using the first loss function value to obtain the first updated model.

[0115] The first updated model is trained using multiple training data points from the training dataset to obtain the third training model;

[0116] Input at least one test data point from the test dataset into the third training model and output the second predicted value of the recovery torque.

[0117] The second loss function value is calculated based on the second predicted recovery torque value and the actual recovery torque value of the test data input into the third training model;

[0118] If the value of the second loss function meets the set error range, then the third training model will be determined as the target recovery torque prediction model.

[0119] In some embodiments, the output module 302 includes:

[0120] The calculation unit is configured to calculate the vehicle's rolling resistance, air resistance, slope resistance, and acceleration resistance based on real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, and real-time road gradient.

[0121] The determining unit is configured to determine the vehicle's target coasting recovery torque based on rolling resistance, air resistance, ramp resistance, and acceleration resistance.

[0122] In some embodiments, the computing unit includes:

[0123] The first calculation component is configured to calculate the rolling resistance of the vehicle based on the real-time vehicle weight and the real-time road friction coefficient.

[0124] The second calculation component is configured to calculate the vehicle's air resistance based on real-time vehicle speed, air resistance coefficient, the vehicle's projected area in the direction of travel, and air density.

[0125] The third calculation component is configured to calculate the vehicle's ramp resistance based on the real-time road gradient and the real-time vehicle weight.

[0126] The fourth calculation component is configured to calculate the vehicle's acceleration drag based on real-time vehicle speed and real-time vehicle weight.

[0127] In some embodiments, the adjustment module 304 includes:

[0128] The adjustment unit is configured to decrease or increase the current coasting recovery torque within a preset time interval according to a preset torque gradient, so that the absolute value of the torque difference between the current coasting recovery torque and the target coasting recovery torque is less than a preset torque threshold.

[0129] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0130] Figure 4 This is a schematic diagram of the structure of a vehicle provided in an embodiment of this application. For ease of description, only the parts related to the embodiment of this application are shown in the figure.

[0131] like Figure 4 As shown, the vehicle includes a coasting energy regeneration control system, which includes a regeneration controller 101 and a drive motor 102; the regeneration controller includes, for example, Figure 3 The vehicle adaptive coasting energy feedback control device shown is shown.

[0132] The technical solution provided in this application embodiment involves the feedback controller 101 collecting real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, real-time road slope, and the current coasting recovery torque of the drive motor 102 when the vehicle is in coasting energy recovery control mode; inputting the real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, and real-time road slope into a preset target recovery torque prediction model, and outputting the target coasting recovery torque; calculating the torque difference between the current coasting recovery torque and the target coasting recovery torque; and adaptively adjusting the current coasting recovery torque of the drive motor 102 if the absolute value of the torque difference is greater than or equal to a preset torque threshold so that the absolute value of the torque difference is less than the preset torque threshold. This application comprehensively considers various factors affecting coasting energy recovery control when the vehicle is in coasting energy recovery control mode (such as vehicle speed, vehicle weight, road friction coefficient, and road gradient), and uses a preset target recovery torque prediction model to predict and obtain the target coasting recovery torque. The target coasting recovery torque obtained in this way is relatively accurate and can help the driver correctly judge the vehicle's deceleration, thereby improving driving safety. Then, by calculating the torque difference between the target coasting recovery torque and the current coasting recovery torque of the drive motor, the current coasting recovery torque of the drive motor is adaptively adjusted according to the torque difference, so that the absolute value of the torque difference is less than a preset torque threshold. This can achieve a smooth transition of power limitation during vehicle shifting while maximizing coasting energy recovery in coasting energy recovery control mode, thereby reducing energy waste.

[0133] Figure 5 This is a schematic diagram of the electronic device 5 provided in an embodiment of this application. Figure 5 As shown, the electronic device 5 of this embodiment includes: a processor 501, a memory 502, and a computer program 503 stored in the memory 502 and executable on the processor 501. When the processor 501 executes the computer program 503, it implements the steps in the various method embodiments described above. Alternatively, when the processor 501 executes the computer program 503, it implements the functions of each module / unit in the various device embodiments described above.

[0134] Electronic device 5 can be a desktop computer, laptop, handheld computer, cloud server, or other electronic device. Electronic device 5 may include, but is not limited to, processor 501 and memory 502. Those skilled in the art will understand that... Figure 5 This is merely an example of electronic device 5 and does not constitute a limitation on electronic device 5. It may include more or fewer components than shown, or different components.

[0135] The processor 501 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0136] The memory 502 can be an internal storage unit of the electronic device 5, such as a hard disk or RAM of the electronic device 5. The memory 502 can also be an external storage device of the electronic device 5, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., equipped on the electronic device 5. The memory 502 can also include both internal and external storage units of the electronic device 5. The memory 502 is used to store computer programs and other programs and data required by the electronic device.

[0137] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0138] If an integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program may include computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately added to or subtracted according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0139] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for adaptive coasting energy feedback control of a vehicle, characterized in that, include: When the vehicle is in coasting energy recovery control mode, the real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, real-time road gradient, and current coasting recovery torque of the drive motor are collected. The real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, and real-time road slope are input into a preset target recovery torque prediction model, and the target coasting recovery torque is output. Calculate the torque difference between the current coasting recovery torque and the target coasting recovery torque; If the absolute value of the torque difference is greater than or equal to a preset torque threshold, the current coasting recovery torque of the drive motor is adaptively adjusted so that the absolute value of the torque difference is less than the preset torque threshold. The target recovery torque prediction model is trained by the following steps: Collect a sample dataset, which includes multiple sample data. Each sample data includes historical vehicle parameters and maximum historical coasting recovery torque when the vehicle or the vehicle and its similar vehicles are in coasting energy feedback control mode. The historical vehicle parameters include historical vehicle speed, historical vehicle weight, historical road friction coefficient, and historical road gradient. The sample dataset is divided into a training dataset and a test dataset. The training dataset includes multiple training data points, and the test dataset includes multiple test data points. Each training data point and each test data point is a single sample data point among the multiple sample data points. The preset initial recovery torque prediction model is trained using multiple training data from the training dataset to obtain the first training model; Input at least one test data point from the test dataset into the first training model and output the first predicted recovery torque value; The first loss function value is calculated based on the first predicted recovery torque value and the actual recovery torque value of the test data input into the first training model; If the value of the first loss function meets the set error range, then the first training model is determined as the target recovery torque prediction model.

2. The method according to claim 1, characterized in that, After calculating the first loss function value based on the first predicted recovery torque value and the actual recovery torque value of the test data input to the first training model, the method further includes: If the first loss function value does not meet the set error range, the first model parameters of the first training model are updated and adjusted using the first loss function value to obtain the first updated model. The test data that was not input into the first training model is used as new training data, or a new training dataset is collected and the first updated model is trained using the new training data to obtain the second training model. If the accuracy of the second training model meets the preset accuracy range, then the second training model is determined as the target recovery torque prediction model.

3. The method according to claim 1, characterized in that, After calculating the first loss function value based on the first predicted recovery torque value and the actual recovery torque value of the test data input to the first training model, the method further includes: If the first loss function value does not meet the set error range, the first model parameters of the first training model are updated and adjusted using the first loss function value to obtain the first updated model. The first updated model is trained using multiple training data points from the training dataset to obtain a third training model; Input at least one test data point from the test dataset into the third training model and output a second predicted recovery torque value. The second loss function value is calculated based on the second predicted recovery torque value and the actual recovery torque value of the test data input into the third training model; If the value of the second loss function meets the set error range, then the third training model is determined as the target recovery torque prediction model.

4. The method according to claim 1, characterized in that, The real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, and real-time road slope are input into a preset target recovery torque prediction model to obtain the target coasting recovery torque, including: Based on the real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, and real-time road gradient, the rolling resistance, air resistance, slope resistance, and acceleration resistance of the vehicle are calculated. The target coasting recovery torque of the vehicle is determined based on the rolling resistance, air resistance, ramp resistance, and acceleration resistance.

5. The method according to claim 4, characterized in that, Based on the real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient, and real-time road gradient, the rolling resistance, air resistance, slope resistance, and acceleration resistance of the vehicle are calculated, including: The rolling resistance of the vehicle is calculated based on the real-time vehicle weight and the real-time road friction coefficient. The air resistance of the vehicle is calculated based on the real-time vehicle speed, air resistance coefficient, projected area of ​​the vehicle in the direction of travel, and air density. The slope resistance of the vehicle is calculated based on the real-time road gradient and the real-time vehicle weight. The acceleration resistance of the vehicle is calculated based on the real-time vehicle speed and real-time vehicle weight.

6. The method according to claim 1, characterized in that, Adaptively adjust the current coasting recovery torque so that the absolute value of the torque difference is less than a preset torque threshold, including: According to a preset torque gradient, the current coasting recovery torque is reduced or increased within a preset time interval, such that the absolute value of the torque difference between the current coasting recovery torque and the target coasting recovery torque is less than a preset torque threshold.

7. A vehicle adaptive coasting energy feedback control device, characterized in that, The apparatus is used to implement the method as described in any one of claims 1 to 6, the apparatus comprising: The data acquisition module is configured to acquire the vehicle's real-time speed, real-time vehicle weight, real-time road friction coefficient, real-time road gradient, and current coasting recovery torque when the vehicle is in coasting energy feedback control mode. The output module is configured to input the real-time vehicle speed, real-time vehicle weight, real-time road friction coefficient and real-time road slope into a preset target recovery torque prediction model, and output the target coasting recovery torque. The calculation module is configured to calculate the torque difference between the current coasting recovery torque and the target coasting recovery torque; The adjustment module is configured to adaptively adjust the current coasting recovery torque if the absolute value of the torque difference is greater than or equal to a preset torque threshold, so that the absolute value of the torque difference is less than the preset torque threshold.

8. A vehicle, characterized in that, It includes a coasting energy feedback control system, which includes a feedback controller and a drive motor; The feedback controller includes the vehicle adaptive coasting energy feedback control device as described in claim 7.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 6.

Images