Commercial vehicle acceleration process dynamic response consistency control method

Abstract

The invention is suitable for the technical field of commercial vehicle driving control, and provides a commercial vehicle acceleration process dynamic response consistency control method, which comprises the following steps: setting expected dynamic response characteristics by designing a reference model, and estimating slow time-varying parameters such as vehicle mass and equivalent resistance coefficient by using an extended Kalman filter; the static gain difference is eliminated through feedforward compensation; meanwhile, a model with an L1 filter is adopted for reference adaptive control to ensure closed-loop dynamic response, and a disturbance observer is used for compensating the influence of transient gradient and wind resistance disturbance. According to the method, the acceleration dynamic response in the acceleration process can be kept consistent under the condition that the mass, the gradient and the wind resistance of the whole vehicle change, and therefore the driving quality is remarkably improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of commercial vehicle driving control, and particularly relates to a commercial vehicle acceleration process dynamic response consistency control method. BACKGROUND

[0002] In recent years, with the rapid development of automobile technology and the continuous enrichment of vehicle models, commercial vehicle users pay more and more attention to vehicle motion quality. Due to the large load of commercial vehicles and the significant difference between empty load and full load, and the influence of factors such as slope and wind resistance during driving, the acceleration response is often inconsistent during acceleration. For example, the vehicle responds agilely when lightly loaded, but the response is obviously lagged when heavily loaded or climbing. This inconsistency of acceleration dynamic response not only affects the driving experience, but also may lead to misoperation, thereby endangering driving safety.

[0003] The Chinese invention patent with the publication number CN113147734B, a closed-loop control method based on driver's longitudinal acceleration intention, acquires the vehicle longitudinal speed, the accelerator pedal opening degree and the actual longitudinal acceleration in real time, combines the stable vehicle speed corresponding pedal opening degree and multi-time domain driving intention, uses Pedal Map to construct the nonlinear mapping of the accelerator pedal and acceleration demand, and realizes the closed-loop control of the vehicle longitudinal motion through PID control. The main innovation of this method is to design the nonlinear mapping relationship between the accelerator pedal and the acceleration demand according to the driver's acceleration intention, and it does not consider the influence of the load multiple level change, slope and wind resistance on the dynamic response of the acceleration process.

[0004] The Chinese invention patent with the publication number CN118775539A, a vehicle starting control method and device based on acceleration, obtains the mass and resistance parameters by identifying the vehicle state, establishes the pedal-acceleration mapping combined with the driver's habit, calculates the required traction force, and dynamically adjusts the engagement position according to the clutch torque transmission characteristic, to realize the precise control of the starting acceleration. Its main innovation is to take acceleration as the core control target, and dynamically calculate the traction force to reduce the dependence on the clutch torque curve. However, this method does not consider the influence of the vehicle itself and the external environment on the acceleration process, and is only applicable to the starting stage. SUMMARY

[0005] The purpose of the embodiment of the application is to provide a commercial vehicle acceleration process dynamic response consistency control method, which aims to solve the problems raised in the above background.

[0006] The embodiment of the application is implemented as follows: a commercial vehicle acceleration process dynamic response consistency control method, comprising the following steps: Step 1: Define the desired dynamic response characteristics. Use a standard second-order system to define the desired acceleration dynamic response characteristics. Obtain the desired dynamic response characteristics by setting the bandwidth and damping. Step 2: Based on the extended Kalman filter, slowly time-varying parameters, including vehicle mass and equivalent drag coefficient, are estimated in real time, and the estimation results are applied to feedforward compensation; Step 3: Calculate and output the feedforward compensation value based on the vehicle mass and equivalent drag coefficient estimated by the extended Kalman filter; Step 4: Estimate the fast time-varying disturbances, including instantaneous slope and sudden change in resistance. When the slope or external resistance changes abruptly, calculate the system's predicted output based on the vehicle driving resistance equation and the dynamic response transfer function of the drive system, and calculate the difference between the predicted output and the actual measured output. After inverting the difference, estimate the transient disturbance force. Step 5: Update control parameters online to make the system transient performance approximate the reference model, while using filters to limit the output bandwidth; Step 6: Adjust the feedforward compensation force Disturbance compensation force and closed-loop adaptive control force The three signals are superimposed to form a comprehensive control signal, which acts on the drive system to control vehicle acceleration. The actual acceleration of the vehicle is compared with the expected acceleration to achieve closed-loop dynamic response control.

[0007] A further technical solution, in step 1, is to use a standard second-order system. The desired acceleration is defined as follows: ; in, For bandwidth, For damping, For Laplace complex variables; Using formula Given a reference acceleration Converted into expected acceleration Output later.

[0008] A further technical solution involves, in step 2, integrating the vehicle's motion state with the parameters to be estimated into a state vector. ,in For vehicle speed, For acceleration, This is the actual driving torque. The angular velocity of the wheel. For the overall vehicle quality, This is the equivalent drag coefficient; The state equation is written as ,in, The derivative of the state vector. represents a state transition function, input of the state equation , is a desired driving torque, expressions of each component are as follows: ; wherein, is a stiffness coefficient, is a total transmission ratio, is a total transmission efficiency, is a wheel rolling radius, is a moment of inertia, is a torque response time constant; is a derivative of the vehicle speed, is a derivative of the acceleration, is a derivative of the actual driving torque, is a derivative of the wheel angular velocity, is a derivative of the vehicle mass, is a derivative of the equivalent resistance coefficient; determining an observation vector , and an observation equation is , is an observation matrix; initialization is performed, and initial state estimation and an initial error covariance matrix are set according to prior information; a prediction calculation is performed on the state vector according to the state equation of the system, and each component is expanded to obtain a linearized Jacobian matrix; a state prediction value and an error covariance prediction value are calculated based on the prediction model; a Kalman gain is calculated according to real-time observation data, and the prediction state is corrected using observation information, and the vehicle mass, the equivalent resistance coefficient and the error covariance matrix are updated online; the posterior state estimation and the error covariance at the current time are taken as initial values at the next time, the prediction and update steps are repeatedly performed, the vehicle mass and the equivalent resistance coefficient are adaptively estimated in real time, and an estimation result is output.

[0009] Further technical solutions, in the step 3, the calculation formula of the feedforward compensation value is as follows: ; wherein, is a vehicle mass estimation value, is an equivalent resistance coefficient estimation value, is a feedforward compensation force.

[0010] Further technical solutions, in the step 4, the estimated disturbance force is subjected to low-pass filtering processing, and acts on a control channel in an opposite sign, transient disturbance is quickly compensated and output, and the calculation formula is as follows: ; In the above formula, is a disturbance compensation force, is a nominal model obtained by solving the automobile running resistance equation and the drive system dynamic response transfer function, is a low-pass filter, is the input of the state equation.

[0011] Further technical solutions, in the step 5, according to the model reference adaptive control principle, define the tracking error , the calculation formula is as follows: ; In the above formula, is the expected acceleration; The prediction error equation is constructed, and the calculation formula is as follows: ; In the above formula, is the prediction error of acceleration, is the estimated value of acceleration, is the true value of acceleration; The gradient type adaptive law is used to update the parameter estimate, and the learning rate and the leakage term are used to adjust the convergence and stability of the parameter update process; The adaptive control quantity is low-pass filtered, the filter bandwidth is set, and the closed-loop adaptive control force is obtained.

[0012] Another purpose of the embodiment of the application is to provide a commercial vehicle acceleration process dynamic response consistency control system based on the above method, comprising: a reference model module, an extended Kalman filter estimation module, a feedforward compensation module, a model reference adaptive control module with a L1 filter, and a disturbance observation and compensation module. The reference model module is used to set the expected dynamic response characteristic, a standard second-order system is used to define the expected acceleration dynamic response characteristic, and the expected dynamic response characteristic is obtained by setting the bandwidth and damping; The extended Kalman filter estimation module is used to estimate the slow time-varying parameters such as the vehicle mass and the equivalent resistance coefficient, and the results are applied to the feedforward compensation to reduce the burden of closed-loop control; The feedforward compensation module calculates the feedforward compensation value based on the vehicle mass and the equivalent resistance coefficient estimated by the extended Kalman filter and outputs it; The disturbance observation and compensation module is used to estimate the instantaneous slope and sudden resistance, when the slope or external resistance suddenly changes, the system prediction output is calculated according to the automobile running resistance equation and the drive system dynamic response transfer function, and the difference between the actual measured output and the system prediction output is calculated, and the transient disturbance force is estimated after being inverted; The model reference adaptive control module with L1 filter is used to update control parameters online, so that the transient performance of the system approximates the reference model, while using the filter to limit the output bandwidth.

[0013] This invention provides a method for consistent dynamic response control during the acceleration process of commercial vehicles. This method sets desired dynamic response characteristics by designing a reference model, estimates slow time-varying parameters such as vehicle mass and equivalent drag coefficient using an extended Kalman filter, and eliminates static gain differences through feedforward compensation. Simultaneously, model reference adaptive control with an L1 filter ensures closed-loop dynamic response, and a disturbance observer compensates for the effects of transient gradient and wind resistance disturbances. This method can maintain consistent acceleration dynamic response during acceleration under varying vehicle mass, gradient, and wind resistance conditions, thereby significantly improving driving quality. Attached Figure Description

[0014] Figure 1 A diagram illustrating the overall control architecture of a dynamic response consistency control method for the acceleration process of a commercial vehicle, provided in an embodiment of the present invention. Figure 2 A diagram illustrating the effect of dynamic response consistency control in accelerating the process. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0016] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0017] like Figure 1 As shown, an embodiment of the present invention provides a dynamic response consistency control method for the acceleration process of a commercial vehicle, comprising the following steps: Step 1: Define the desired dynamic response characteristics. Use a standard second-order system to define the desired acceleration dynamic response characteristics. Obtain the desired dynamic response characteristics by setting the bandwidth and damping.

[0018] Through the standard second-order system The desired acceleration is defined as follows: ; in, For bandwidth, For damping, It is a Laplace complex variable.

[0019] Using formula Given a reference acceleration Converted into expected acceleration rear output.

[0020] Step 2: Real-time estimation of slow time-varying parameters including vehicle mass and equivalent drag coefficient based on extended Kalman filter, and application of the estimation results to feedforward compensation to reduce the burden of closed-loop control.

[0021] Integrate vehicle motion state and parameters to be estimated into state vector wherein is vehicle speed, is acceleration, is actual driving torque, is wheel angular velocity, is vehicle mass, is equivalent drag coefficient.

[0022] Write the state equation as wherein, denotes the derivative of the state vector, denotes the state transition function, the input of the state equation , is the desired driving torque, and the expressions of the components are as follows: ; wherein, is the stiffness coefficient, is the total transmission ratio, is the total transmission efficiency, is the wheel rolling radius, is the moment of inertia, is the torque response time constant; is the derivative of the vehicle speed, is the derivative of the acceleration, is the derivative of the actual driving torque, is the derivative of the wheel angular velocity, is the derivative of the vehicle mass, is the derivative of the equivalent drag coefficient.

[0023] Determine the observation vector , and the observation equation is , is the observation matrix.

[0024] Initialization: Set the initial state estimate and the initial error covariance matrix according to the prior information.

[0025] According to the state equation of the system, the state vector is predicted and calculated, and the linearized Jacobian matrix is obtained by expanding the components; the state prediction value and the error covariance prediction value are calculated based on the prediction model.

[0026] The Kalman gain is calculated according to real-time observation data, and the predicted state is corrected by using observation information to update the vehicle mass, equivalent resistance coefficient and error covariance matrix online.

[0027] The posterior state estimation value and error covariance at the current time are taken as initial values at the next time, and the prediction and update steps are repeatedly performed to perform real-time adaptive estimation on the vehicle mass and equivalent resistance coefficient and output the estimation results.

[0028] Step 3: Calculate the feedforward compensation value based on the vehicle mass and equivalent resistance coefficient estimated by the extended Kalman filter and output, and the calculation formula of the feedforward compensation value is as follows: ; Among them, is the vehicle mass estimation value, is the equivalent resistance coefficient estimation value, is the feedforward compensation force.

[0029] Step 4: Estimate the instantaneous slope, sudden resistance and other fast time-varying disturbances. When the slope or external resistance suddenly changes, calculate the system predicted output according to the automobile running resistance equation and the driving system dynamic response transfer function, and calculate the difference with the actual measured output. After inversion, the instantaneous disturbance force is estimated.

[0030] The estimated disturbance force is low-pass filtered and acts on the control channel in the opposite direction. The transient disturbance is quickly compensated and output, and the calculation formula is as follows: ; In the above formula, is the disturbance compensation force, is the nominal model obtained by simultaneously solving the automobile running resistance equation and the driving system dynamic response transfer function, is a low-pass filter, is the input of the state equation.

[0031] Step 5: Update the control parameters online to make the system transient performance approach the reference model, and at the same time use the filter to limit the output bandwidth.

[0032] According to the model reference adaptive control principle, define the tracking error , and the calculation formula is as follows: ; In the above formula, is the expected acceleration.

[0033] Build a prediction error equation, and the calculation formula is as follows: ; In the above formula, is the prediction error of acceleration, is an estimated value of acceleration, is a true value of acceleration.

[0034] The parameter estimated value is updated by using a gradient type adaptive law, and a learning rate and a leakage term are used to adjust the convergence and stability of the parameter updating process.

[0035] The adaptive control quantity is low-pass filtered, a filter bandwidth is set, the accuracy of the low and medium frequency adaptive response is ensured, and high frequency interference is suppressed, so as to obtain a closed-loop adaptive control force .

[0036] Step 6: superimposing , and three items to form a comprehensive control signal, which acts on the driving system to control the vehicle acceleration, compares the actual acceleration of the vehicle with the expected acceleration, and realizes closed-loop dynamic response control.

[0037] As shown in Figure 1 , a commercial vehicle acceleration process dynamic response consistency control system provided by another embodiment of the application is based on the above method and comprises a reference model module, an extended Kalman filter estimation module, a feedforward compensation module, a model reference adaptive control module with an L1 filter, and a disturbance observation and compensation module. The reference model module is used to set the expected dynamic response characteristic, a standard second-order system is used to define the expected acceleration dynamic response characteristic, and the expected dynamic response characteristic is obtained by setting the bandwidth and damping; The extended Kalman filter estimation module is used to estimate slow time-varying parameters such as the vehicle mass and the equivalent resistance coefficient, and the results are applied to feedforward compensation to reduce the burden of closed-loop control. The feedforward compensation module calculates and outputs the feedforward compensation value based on the vehicle mass and the equivalent resistance coefficient estimated by the extended Kalman filter. The disturbance observation and compensation module is used to estimate fast time-varying disturbances such as instantaneous slope and sudden resistance, when the slope or external resistance suddenly changes, the system predicted output is calculated according to the automobile driving resistance equation and the driving system dynamic response transfer function, and the difference between the actual measured output and the predicted output is calculated, and the transient disturbance force is estimated after inversion. The model reference adaptive control module with an L1 filter is used to update the control parameters online, so that the system transient performance approximates the reference model, and the filter is used to limit the output bandwidth.

[0038] As a preferred embodiment of the present application, the above method is used to take a certain commercial vehicle as the research object, two different working conditions are selected to compare the control effect of acceleration dynamic response consistency, working condition one is that the vehicle mass is 40000 kg, the equivalent resistance coefficient is 300 Ns / m, and the initial vehicle speed is 20 m / s; working condition two is that the vehicle mass is 30000 kg, the equivalent resistance coefficient is 500 Ns / m, and the vehicle speed is 10 m / s. It can be known from the comparison that the method has good control effect. Figure 2

[0039] The above only describes preferred embodiments of the present application and is not used to limit the present application, and any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.​

Claims

1. A method for consistent dynamic response control during the acceleration process of a commercial vehicle, characterized in that, Includes the following steps: Step 1: Define the desired dynamic response characteristics. Use a standard second-order system to define the desired acceleration dynamic response characteristics. Obtain the desired dynamic response characteristics by setting the bandwidth and damping. Step 2: Based on the extended Kalman filter, slowly time-varying parameters, including vehicle mass and equivalent drag coefficient, are estimated in real time, and the estimation results are applied to feedforward compensation; Step 3: Calculate and output the feedforward compensation value based on the vehicle mass and equivalent drag coefficient estimated by the extended Kalman filter; Step 4: Estimate the fast time-varying disturbances, including instantaneous slope and sudden change in resistance. When the slope or external resistance changes abruptly, calculate the system's predicted output based on the vehicle driving resistance equation and the dynamic response transfer function of the drive system, and calculate the difference between the predicted output and the actual measured output. After inverting the difference, estimate the transient disturbance force. Step 5: Update control parameters online to make the system transient performance approximate the reference model, while using filters to limit the output bandwidth; Step 6: Adjust the feedforward compensation force Disturbance compensation force and closed-loop adaptive control force The three signals are superimposed to form a comprehensive control signal, which acts on the drive system to control vehicle acceleration. The actual acceleration of the vehicle is compared with the expected acceleration to achieve closed-loop dynamic response control.

2. The method for consistent dynamic response control during acceleration of commercial vehicles according to claim 1, characterized in that, Through the standard second-order system The desired acceleration is defined as follows: ； in, For bandwidth, For damping, For Laplace complex variables; Using formula Given a reference acceleration Converted into expected acceleration Output later.

3. The method for consistent dynamic response control during acceleration of commercial vehicles according to claim 2, characterized in that, In step 2, the vehicle motion state and the parameters to be estimated are integrated into a state vector. ,in For vehicle speed, For acceleration, This is the actual driving torque. The angular velocity of the wheel. For the overall vehicle quality, This is the equivalent drag coefficient; The state equation is written as ,in, The derivative of the state vector. Represents the state transition function, the input to the state equation. , The expressions for each component of the desired driving torque are as follows: ； in, This is the stiffness coefficient. The total transmission ratio is... The total transmission efficiency. The radius of the wheel's rolling motion. For rotational inertia, This is the torque response time constant; The derivative of the vehicle speed. The derivative of acceleration, This is the derivative of the actual driving torque. The derivative of the wheel's angular velocity. Let be the derivative of the total vehicle mass. The derivative of the equivalent drag coefficient; Determine the observation vector The observation equation is , The observation matrix; Initialization is performed, and the initial state estimate and initial error covariance matrix are set based on prior information; The state vector is predicted and calculated based on the system's state equation, and the linearized Jacobian matrix is ​​obtained by expanding each component; the predicted state value and the predicted error covariance value are calculated based on the prediction model. The Kalman gain is calculated based on real-time observation data, and the predicted state is corrected using the observation information. The vehicle mass, equivalent drag coefficient, and error covariance matrix are updated online. Using the posterior state estimate and error covariance at the current moment as the initial value for the next moment, the prediction and update steps are repeated to perform real-time adaptive estimation of vehicle mass and equivalent drag coefficient and output the estimation results.

4. The method for consistent dynamic response control during acceleration of commercial vehicles according to claim 3, characterized in that, In step 3, the formula for calculating the feedforward compensation value is as follows: ； in, This is an estimated value for the vehicle's total weight. This is an estimate of the equivalent drag coefficient. It is a feedforward compensation force.

5. The method for consistent dynamic response control during acceleration of commercial vehicles according to claim 4, characterized in that, In step 4, the estimated disturbance force is low-pass filtered and applied to the control channel with the opposite sign to quickly compensate for the transient disturbance and output the result. The calculation formula is as follows: ； In the above formula, For disturbance compensation force, The nominal model is obtained by simultaneously solving the equations for vehicle driving resistance and the transfer function of the dynamic response of the drive system. It is a low-pass filter. This is the input to the state equation.

6. The method for consistent dynamic response control during acceleration of commercial vehicles according to claim 5, characterized in that, In step 5, the tracking error is defined according to the principle of model reference adaptive control. The calculation formula is as follows: ； In the above formula, For the desired acceleration; The prediction error equation is constructed, and the calculation formula is as follows: ； In the above formula, For the prediction error of acceleration, This is an estimate of the acceleration. This represents the true value of the acceleration. Gradient-based adaptive law is used to update parameter estimates, and the convergence and stability of the parameter update process are adjusted using the learning rate and leakage term. The adaptive control input is low-pass filtered, and the filter bandwidth is set to obtain the closed-loop adaptive control force. .

Images