阅读说明：本技术 一种线控转向系统及故障诊断方法 (Wire-controlled steering system and fault diagnosis method ) 是由 陈锋 傅直全 胡斐 俞碧君 于 2021-07-20 设计创作，主要内容包括：本发明公开了一种线控转向系统及故障诊断方法,包括转向盘模块、转向执行模块、路感反馈模块、ECU控制模块、转向盘转角传感器、车速传感器、横摆角速度传感器。且本方案的故障诊断方法完全基于车辆动力学模型的前轮转角估计算法相比,基于双向长短时记忆网络可以不依赖于精确的动力学模型,可以很好地对高度复杂性、非线性以及强耦合性的系统状态进行估计；相比于分析一次测量值的传感器故障诊断方法,基于双层状态机分别对前轮转角信号的范围、信号差值进行累计诊断可以减小对传感器状态的误判,提高诊断结果的可靠性能。(The invention discloses a line control steering system and a fault diagnosis method, which comprise a steering wheel module, a steering execution module, a road feel feedback module, an ECU control module, a steering wheel corner sensor, a vehicle speed sensor and a yaw rate sensor. Compared with a front wheel steering angle estimation calculation method based on a vehicle dynamic model, the fault diagnosis method based on the scheme has the advantages that the method is based on a bidirectional long-time and short-time memory network, does not depend on an accurate dynamic model, and can well estimate the system state with high complexity, nonlinearity and strong coupling; compared with a sensor fault diagnosis method for analyzing a measured value once, accumulated diagnosis is respectively carried out on the range and the signal difference value of the front wheel steering angle signal based on the double-layer state machine, so that misjudgment on the state of the sensor can be reduced, and the reliability of a diagnosis result is improved.)

1. A wire-controlled steering system is characterized by comprising a steering wheel module, a steering execution module, a road feel feedback module, an ECU control module, a steering wheel corner sensor, a vehicle speed sensor and a yaw rate sensor;

the steering execution module comprises a steering motor, a speed reducer, a gear, a rack, a steering tie rod and wheels; the output shaft of the steering motor is connected with a gear through a worm gear reducer, the gear is meshed with a rack, and the rack drives wheels through a steering tie rod;

the road sense feedback module comprises a road sense motor and a road sense motor reducer, and an output shaft of the road sense motor is connected with the lower end of the steering column through the road sense motor reducer and used for transmitting road sense to the steering wheel;

the steering wheel corner sensor is arranged on the steering column and used for measuring the steering wheel corner and transmitting the steering wheel corner to the ECU control module;

the vehicle speed sensor is arranged on a wheel and used for acquiring the longitudinal vehicle speed of the automobile and transmitting the longitudinal vehicle speed to the ECU control module;

the yaw rate sensors are all arranged at the mass center of the frame and used for acquiring the yaw rate of the automobile and transmitting the yaw rate to the ECU control module;

and the ECU control module controls the work of the road sensing motor and the steering motor according to the received steering wheel angle signal, the longitudinal vehicle speed signal and the yaw rate signal.

2. A method for diagnosing a malfunction of a front wheel steering sensor of a steer-by-wire vehicle, comprising the steer-by-wire system of claim 1, comprising the steps of:

step 2.1), a front wheel steering angle sensor measures to obtain a front wheel steering angle signal s1, an extended Kalman filtering algorithm estimates to obtain a front wheel steering angle signal s2, and a bidirectional long-short time memory network estimates to obtain a front wheel steering angle signal s 3;

step 2.2), the angle range detection module respectively performs accumulation detection on the ranges of the front wheel steering angle signals s1, s2 and s3 by using a state machine, if the front wheel steering angle signals s1, s2 and s3 exceed a set threshold theta which is 40 degrees, the value of a corresponding counter i is i +1, otherwise, the counter i is the original value; when the value of the counter is greater than the set threshold, the output flag1 is 0, the flag2 is 0, and the flag3 is 0, otherwise, the flag1 is 1, the flag2 is 1, and the flag3 is 1;

step 2.3), when the flag1 is equal to 1, the flag2 is equal to 1, and the flag3 is equal to 1, it is indicated that all front wheel steering angle signals are in a reasonable range, and the angle range detection module outputs front wheel steering angle signals s1' ═ s1, s2' ═ s2, and s3' ═ s3 respectively; when the flag1 is equal to 0, the flag2 is equal to 0, and the flag3 is equal to 0, the front wheel steering angle signal is out of a reasonable range;

step 2.4), the angle deviation detection module respectively performs accumulative detection on the angle deviation signals s11 ═ s1'-s2' |, s22 ═ s1'-s3' |, and s33 ═ s2'-s3' |, if the angle deviation signals exceed a set threshold, the value of the corresponding counter i is i ═ i +1, otherwise, the counter i is the original value; when the value of the counter is greater than the set threshold, the output flag11 is 0, the flag22 is 0, and the flag33 is 0, otherwise, the flag11 is 1, the flag22 is 1, and the flag33 is 1;

step 2.5), the fault diagnosis module respectively performs logic or operation on < flag11 >, flag22>, < flag11 >, flag33>, < flag22 and flag33>, and judges the fault condition of each signal according to the fault diagnosis table.

3. The steer-by-wire vehicle front wheel steering angle sensor malfunction diagnosis method according to claim 2, characterized in that the front wheel steering angle signal sets the threshold value at theta-40 °.

4. The steer-by-wire vehicle front wheel steering angle sensor malfunction diagnosis method according to claim 2, characterized in that the angle deviation signal threshold value set by the angle deviation detection module is diff-0.2 °.

5. The steer-by-wire vehicle front wheel steering angle sensor malfunction diagnosis method according to claim 2, characterized in that the threshold values of the counters in step 2.2) and step 2.4) are both 5.

6. The method for diagnosing the fault of the front wheel steering sensor of the steer-by-wire vehicle according to claim 2, wherein the step 2.1) method for estimating the front wheel steering angle signal by the extended kalman filter algorithm is as follows:

step 3.1), a three-degree-of-freedom model of the vehicle is established, and front wheel steering angle and first-order derivative thereof, yaw angular velocity, mass center slip angle and longitudinal vehicle speed are selected as state variablesThe input quantity is longitudinal acceleration and steering motor current [ a ]_x,I]The observation vector is the lateral acceleration [ a ]_y]：

In the formula, a is the distance from the mass center of the automobile to the front axle; b is the distance from the mass center of the automobile to the rear axle; v. of_xIs the longitudinal speed of the vehicle; delta is a front wheel corner; k is a radical of₁Front wheel cornering stiffness; k is a radical of₂Is rear wheel cornering stiffness; alpha is alpha_fIs a front wheel side slip angle; alpha is alpha_rIs a rear wheel side slip angle; m is the mass of the whole vehicle; beta is the vehicle body mass center slip angle; gamma is a yaw angular velocity; a is_yIs the longitudinal acceleration of the vehicle; i is_zThe moment of inertia of the automobile around the z axis; t is t_pA tire drag distance; t is t_mThe main pin is inwards inclined; m_rThe mass of the rack; b is_rThe rack damping coefficient; k_tIs the torque coefficient of the steering motor; n is the reduction ratio of the steering motor reducer; i is the motor current of the steering system; eta is the motor efficiency; r is_pThe radius of the pinion.

Step 3.2), carrying out linearization processing on the nonlinear state equation and the measurement equation of the system, carrying out Taylor series expansion on the state equation and the measurement equation, and respectively obtaining a Jacobian matrix corresponding to the state equation and the measurement equation according to the formula (1) and the formula (2):

step 3.3), solving the state transition matrix phi of the system_kAnd estimating a front wheel steering angle signal according to the input quantity and the observation vector:

Φ_k＝I_5×5+F_k·T_s (5)

in the formula I_5x5Is an identity matrix; t is_sIs the sampling time.

7. The method for diagnosing the fault of the front wheel steering sensor of the steer-by-wire vehicle according to claim 2, wherein the step 2.1) of bidirectional long-and-short time memory network method for estimating the front wheel steering angle signal is as follows:

step 4.1), collecting a yaw velocity signal gamma, a mass center side deviation angle signal beta and a steering wheel angle signal theta of the vehicle_swVehicle speed signal v_xAnd carrying out normalization treatment on the obtained product:

in the formula, X is data needing normalization; x_minIs the minimum value of the data; x_maxIs the maximum value of the data; x is normalized data;

step 4.2), setting the number of network layers to be 3, the number of neurons in each layer to be 50, the activation function to be sigmoid, the batch size to be 400, the training times to be 2000 and the optimizer to be a network model of RMSprop;

step 4.3), taking 70% of the normalized data as a training set, taking the rest data as a test set to carry out model training and testing, and selecting a root mean square error RMSE as an evaluation index of the model precision:

and 4.4), if the RMSE is larger than a preset precision threshold value, skipping to the step 4.2) and modifying the model parameters, and if not, directly using the trained model for front wheel steering angle estimation.

Technical Field

The invention relates to the field of automobile steer-by-wire, in particular to a steer-by-wire system and a fault diagnosis method.

Background

Redundant control of the sensors that steer the vehicle must be provided in order to improve the active safety of the vehicle. Redundancy control includes hardware redundancy and analytical redundancy. Sensor hardware redundancy is mainly achieved by increasing the number of physical sensors, and generally speaking, the more vehicle state sensors, the more redundant and robust the system. However, some vehicle state sensors are expensive or cannot be used in mass production on vehicles due to technical constraints (reliability, accuracy, stability, etc.). Compared to sensor hardware redundancy, sensor analytical redundancy techniques can eliminate physical sensors without affecting system reliability, which greatly reduces overall system cost.

However, the current commonly used analytical redundancy is based on vehicle dynamics model for state estimation, but in practice most systems exhibit high complexity, nonlinearity and coupling, which makes it difficult to build an accurate system model. The inaccuracy of the model and the randomness of the observation noise may cause serious problems such as reduced accuracy of the result, filter divergence and even failure to estimate. In addition, the analytical easy method based on the triplex redundancy is mostly to directly compare the difference values of three signals so as to judge the fault state of the sensor. However, the method based on one-time judgment is easy to generate misjudgment on the actual state of the sensor due to the existence of external transient interference, measurement noise and the like.

Disclosure of Invention

The technical problem to be solved by the invention is to provide a wire-controlled steering system and a fault diagnosis method aiming at the defects involved in the background technology.

The invention adopts the following technical scheme for solving the technical problems:

a wire-controlled steering system comprises a steering wheel module, a steering execution module, a road feel feedback module, an ECU control module, a steering wheel corner sensor, a vehicle speed sensor and a yaw rate sensor;

the steering wheel module comprises a steering wheel and a steering column, and the upper end of the steering column is fixedly connected with the steering wheel;

the steering execution module comprises a steering motor, a worm and gear reducer, a gear, a rack, a steering tie rod and wheels; the output shaft of the steering motor is connected with a gear through a worm gear reducer, the gear is meshed with a rack, and the rack drives wheels through a steering tie rod;

the road sense feedback module comprises a road sense motor and a road sense motor reducer, and an output shaft of the road sense motor is connected with the lower end of the steering column through the road sense motor reducer and used for transmitting road sense to the steering wheel;

the steering wheel corner sensor is arranged on the steering column and used for measuring the steering wheel corner and transmitting the steering wheel corner to the ECU control module;

the vehicle speed sensor is arranged on a wheel and used for acquiring the longitudinal vehicle speed of the automobile and transmitting the longitudinal vehicle speed to the ECU control module;

the yaw rate sensors are all arranged at the mass center of the frame and used for acquiring the yaw rate of the automobile and transmitting the yaw rate to the ECU control module;

and the ECU control module controls the work of the road sensing motor and the steering motor according to the received steering wheel angle signal, the longitudinal vehicle speed signal and the yaw rate signal.

The invention also discloses a method for diagnosing the fault of the front wheel steering angle sensor of the steer-by-wire vehicle, which comprises the following steps:

step 2.1), a front wheel steering angle sensor measures to obtain a front wheel steering angle signal s1, an extended Kalman filtering algorithm estimates to obtain a front wheel steering angle signal s2, and a bidirectional long-short time memory network estimates to obtain a front wheel steering angle signal s 3;

step 2.2), the angle range detection module respectively performs accumulation detection on the ranges of the front wheel steering angle signals s1, s2 and s3 by using a state machine, if the front wheel steering angle signals s1, s2 and s3 exceed a set threshold theta which is 40 degrees, the value of a corresponding counter i is i +1, otherwise, the counter i is the original value; when the value of the counter is greater than the set threshold value N, the output flag1 is 0, the flag2 is 0, and the flag3 is 0, otherwise, the flag1 is 1, the flag2 is 1, and the flag3 is 1;

step 2.3), when the flag1 is equal to 1, the flag2 is equal to 1, and the flag3 is equal to 1, it is indicated that all front wheel steering angle signals are in a reasonable range, and the angle range detection module outputs front wheel steering angle signals s1' ═ s1, s2' ═ s2, and s3' ═ s3 respectively; when the flag1, the flag2 and the flag3 are respectively equal to 0, the front wheel steering angle signal is out of a reasonable range, and the angle range detection module respectively outputs the front wheel steering angle signal s1, s2, 80 and s3, 90 degrees;

step 2.4), the angle deviation detection module respectively performs accumulative detection on the angle deviation signals s11 ═ s1'-s2' |, s22 ═ s1'-s3' |, and s33 ═ s2'-s3' |, if the angle deviation signals exceed a set threshold diff ═ 0.2 °, the value of the corresponding counter i is i ═ i +1, otherwise, the counter i is the original value; when the value of the counter is greater than the set threshold value N, the output flag11 is 0, the flag22 is 0, and the flag33 is 0, otherwise, the flag11 is 1, the flag22 is 1, and the flag33 is 1;

step 2.5), the fault diagnosis module respectively performs logic or operation on < flag11 >, flag22>, < flag11 >, flag33>, < flag22 and flag33>, and judges the fault condition of each signal according to the fault diagnosis table;

TABLE 1 Fault diagnosis Table

The method for diagnosing the fault of the front wheel steering angle sensor of the steer-by-wire vehicle comprises the step 2.1) that the method for estimating the front wheel steering angle signal by the extended Kalman filtering algorithm comprises the following steps:

step 3.1), a three-degree-of-freedom model of the vehicle is established, and front wheel steering angle and first-order derivative thereof, yaw angular velocity, mass center slip angle and longitudinal vehicle speed are selected as state variablesThe input quantity is longitudinal acceleration and steering motor current [ a ]_x,I]The observation vector is the lateral acceleration [ a ]_y]：

In which a is a carThe distance of the center of mass to the front axis; b is the distance from the mass center of the automobile to the rear axle; v. of_xIs the longitudinal speed of the vehicle; delta is a front wheel corner; k is a radical of₁Front wheel cornering stiffness; k is a radical of₂Is rear wheel cornering stiffness; alpha is alpha_fIs a front wheel side slip angle; alpha is alpha_rIs a rear wheel side slip angle; m is the mass of the whole vehicle; beta is the vehicle body mass center slip angle; gamma is a yaw angular velocity; a is_yIs the longitudinal acceleration of the vehicle; i is_zThe moment of inertia of the automobile around the z axis; t is t_pA tire drag distance; t is t_mThe main pin is inwards inclined; m_rThe mass of the rack; b is_rThe rack damping coefficient; k_tIs the torque coefficient of the steering motor; n is the reduction ratio of the steering motor reducer; i is the motor current of the steering system; eta is the motor efficiency; r is_pThe radius of the pinion.

Step 3.2), carrying out linearization processing on the nonlinear state equation and the measurement equation of the system, carrying out Taylor series expansion on the state equation and the measurement equation, and respectively obtaining a Jacobian matrix corresponding to the state equation and the measurement equation according to the formula (1) and the formula (2):

step 3.3), solving the state transition matrix phi of the system_kAnd estimating a front wheel steering angle signal according to the input quantity and the observation vector:

Φ_k＝I_5×5+F_k·T_s (5)

in the formula I_5x5Is an identity matrix; t is_sIs the sampling time.

As the fault diagnosis method for the front wheel steering angle sensor of the steer-by-wire vehicle, the method for estimating the front wheel steering angle signal by the bidirectional long-time and short-time memory network in the step 2.1) is as follows:

step 4.1), collecting yaw angular velocity signals gamma and mass center lateral deviation angle signals of the vehicleNumber beta, steering wheel angle signal theta_swVehicle speed signal v_xAnd carrying out normalization treatment on the obtained product:

in the formula, X is data needing normalization; x_minIs the minimum value of the data; x_maxIs the maximum value of the data; x is normalized data;

step 4.2), setting the number of network layers to be 3, the number of neurons in each layer to be 50, the activation function to be sigmoid, the batch size to be 400, the training times to be 2000 and the optimizer to be a network model of RMSprop;

step 4.3), taking 70% of the normalized data as a training set, taking the rest data as a test set to carry out model training and testing, and selecting a root mean square error RMSE as an evaluation index of the model precision:

and 4.4), if the RMSE is larger than a preset precision threshold value, skipping to the step 4.2) and modifying the model parameters, and if not, directly using the trained model for front wheel steering angle estimation.

Compared with the prior art, the invention adopting the technical scheme has the following technical effects:

compared with a front wheel steering angle estimation method completely based on a vehicle dynamic model, the method based on the bidirectional long-time and short-time memory network can well estimate the system state with high complexity, nonlinearity and strong coupling without depending on an accurate dynamic model; compared with a sensor fault diagnosis method for analyzing a measured value once, accumulated diagnosis is respectively carried out on the range and the signal difference value of the front wheel steering angle signal based on the double-layer state machine, so that misjudgment on the state of the sensor can be reduced, and the reliability of a diagnosis result is improved.

Drawings

FIG. 1 is a schematic diagram of the fault diagnosis of a steer-by-wire vehicle front wheel steering angle sensor of the present invention.

FIG. 2 is a state machine based angular range accumulation diagnostic schematic of the present invention.

FIG. 3 is a schematic diagram of the state machine based angular deviation accumulation diagnostic of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples.

Example 1

The technical scheme of the invention is further explained in detail by combining the attached drawings:

the present invention may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. In the drawings, components are exaggerated for clarity.

The invention discloses a line control steering system, which comprises a steering wheel module, a steering execution module, a road feel feedback module, an ECU control module, a steering wheel corner sensor, a vehicle speed sensor and a yaw rate sensor, wherein the steering wheel module is connected with the road feel feedback module;

the steering wheel module comprises a steering wheel and a steering column, and the upper end of the steering column is fixedly connected with the steering wheel;

the steering execution module comprises a steering motor, a worm and gear reducer, a gear, a rack, a steering tie rod and wheels; the output shaft of the steering motor is connected with a gear through a worm gear reducer, the gear is meshed with a rack, and the rack drives wheels through a steering tie rod;

the road sense feedback module comprises a road sense motor and a road sense motor reducer, and an output shaft of the road sense motor is connected with the lower end of the steering column through the road sense motor reducer and used for transmitting road sense to the steering wheel;

the steering wheel corner sensor is arranged on the steering column and used for measuring the steering wheel corner and transmitting the steering wheel corner to the ECU control module;

the vehicle speed sensor is arranged on a wheel and used for acquiring the longitudinal vehicle speed of the automobile and transmitting the longitudinal vehicle speed to the ECU control module;

the yaw rate sensors are all arranged at the mass center of the frame and used for acquiring the yaw rate of the automobile and transmitting the yaw rate to the ECU control module;

and the ECU control module controls the work of the road sensing motor and the steering motor according to the received steering wheel angle signal, the longitudinal vehicle speed signal and the yaw rate signal.

As shown in fig. 1, the invention also discloses a method for diagnosing the fault of the front wheel steering angle sensor of the steer-by-wire vehicle, which comprises the following steps:

step 2.1), a front wheel steering angle sensor measures to obtain a front wheel steering angle signal s1, an extended Kalman filtering algorithm estimates to obtain a front wheel steering angle signal s2, and a bidirectional long-short time memory network estimates to obtain a front wheel steering angle signal s 3;

step 2.2), as shown in fig. 2, the angle range detection module performs accumulation detection on the ranges of the front wheel steering angle signals s1, s2, and s3 by using a state machine, if the front wheel steering angle signals s1, s2, and s3 exceed the set threshold value theta equal to 40 °, the value of the corresponding counter i is i +1, otherwise, the counter i is the original value; when the value of the counter is greater than the set threshold value N, the output flag1 is 0, the flag2 is 0, and the flag3 is 0, otherwise, the flag1 is 1, the flag2 is 1, and the flag3 is 1;

step 2.3), when the flag1 is equal to 1, the flag2 is equal to 1, and the flag3 is equal to 1, it is indicated that all front wheel steering angle signals are in a reasonable range, and the angle range detection module outputs front wheel steering angle signals s1' ═ s1, s2' ═ s2, and s3' ═ s3 respectively; when the flag1, the flag2 and the flag3 are respectively equal to 0, the front wheel steering angle signal is out of a reasonable range, and the angle range detection module respectively outputs the front wheel steering angle signal s1, s2, 80 and s3, 90 degrees;

step 2.4), as shown in fig. 3, the angle deviation detecting module respectively performs cumulative detection on the angle deviation signals s11 ═ s1'-s2' |, s22 ═ s1'-s3' |, s33 ═ s2'-s3' |, if the angle deviation signals exceed the set threshold diff ═ 0.2 °, the value of the corresponding counter i is i ═ i +1, otherwise, the counter i is the original value; when the value of the counter is greater than the set threshold value N, the output flag11 is 0, the flag22 is 0, and the flag33 is 0, otherwise, the flag11 is 1, the flag22 is 1, and the flag33 is 1;

step 2.5), the fault diagnosis module respectively performs logic or operation on < flag11 >, flag22>, < flag11 >, flag33>, < flag22 and flag33>, and judges the fault condition of each signal according to the fault diagnosis table;

TABLE 1 Fault diagnosis Table

As the fault diagnosis method for the front wheel steering angle sensor of the steer-by-wire vehicle, the method for estimating the front wheel steering angle signal by the extended Kalman filtering algorithm in the step 2.1) is as follows:

step 3.1), a three-degree-of-freedom model of the vehicle is established, and front wheel steering angle and first-order derivative thereof, yaw angular velocity, mass center slip angle and longitudinal vehicle speed are selected as state variablesThe input quantity is longitudinal acceleration and steering motor current [ a ]_x,I]The observation vector is the lateral acceleration [ a ]_y]：

In the formula, a is the distance from the mass center of the automobile to the front axle; b is the distance from the mass center of the automobile to the rear axle; v. of_xIs the longitudinal speed of the vehicle; delta is a front wheel corner; k is a radical of₁Front wheel cornering stiffness; k is a radical of₂Is rear wheel cornering stiffness; alpha is alpha_fIs a front wheel side slip angle; alpha is alpha_rIs a rear wheel side slip angle; m is the mass of the whole vehicle; beta is the vehicle body mass center slip angle; gamma is a yaw angular velocity; a is_yIs the longitudinal acceleration of the vehicle; i is_zThe moment of inertia of the automobile around the z axis; t is t_pA tire drag distance; t is t_mThe main pin is inwards inclined;M_rthe mass of the rack; b is_rThe rack damping coefficient; k_tIs the torque coefficient of the steering motor; n is the reduction ratio of the steering motor reducer; i is the motor current of the steering system; eta is the motor efficiency; r is_pThe radius of the pinion.

Step 3.2), carrying out linearization processing on the nonlinear state equation and the measurement equation of the system, carrying out Taylor series expansion on the state equation and the measurement equation, and respectively obtaining a Jacobian matrix corresponding to the state equation and the measurement equation according to the formula (1) and the formula (2):

step 3.3), solving the state transition matrix phi of the system_kAnd estimating a front wheel steering angle signal according to the input quantity and the observation vector:

Φ_k＝I_5×5+F_k·T_s (5)

in the formula I_5x5Is an identity matrix; t is_sIs the sampling time.

As the fault diagnosis method for the front wheel steering angle sensor of the steer-by-wire vehicle, the method for estimating the front wheel steering angle signal by the bidirectional long-time and short-time memory network in the step 2.1) is as follows:

step 4.1), collecting a yaw velocity signal gamma, a mass center side deviation angle signal beta and a steering wheel angle signal theta of the vehicle_swVehicle speed signal v_xAnd carrying out normalization treatment on the obtained product:

in the formula, X is data needing normalization; x_minIs the minimum value of the data; x_maxIs the maximum value of the data; x is normalized data;

step 4.2), setting the number of network layers to be 3, the number of neurons in each layer to be 50, the activation function to be sigmoid, the batch size to be 400, the training times to be 2000 and the optimizer to be a network model of RMSprop;

step 4.3), taking 70% of the normalized data as a training set, taking the rest data as a test set to carry out model training and testing, and selecting a root mean square error RMSE as an evaluation index of the model precision:

and 4.4), if the RMSE is larger than a preset precision threshold value, skipping to the step 4.2) and modifying the model parameters, and if not, directly using the trained model for front wheel steering angle estimation.

While the invention has been described in terms of its preferred embodiments, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.