Research On Steering Torque Control Of Intelligent Vehicle Based On Driver Intention Recognition

With the rapid development of automobile technology,driverless cars will be industrialized and intelligent gradually.Therefore,driverless cars have gradually become the focus of many scholars around the world.Currently,unmanned driving is usually controlled based on the steering wheel's angle,which often makes it impossible for the driver to meddle steering,thus completely breaking away from the automatic driving control loop,easily resulting in the reduction of "situational awareness" and other adverse factors.In order to solve many problems faced by unmanned driving,many scholars at home and abroad have put forward the concept of "human-machine co-driving" intelligent vehicle driving,humanmachine joint participation in the control loop of automatic driving.As human and machine participate in the control loop,it is very important to identify the driver's intention,otherwise there will be making conflict between human and machine's decision.In order to identify the driver's intention,the markov theory is introduced in detail,and a two-layer hidden markov model with high identification accuracy is proposed to identify the driver's intention.In order to ensure the accuracy of driver intention recognition,t-text algorithm was used to eliminate the outliers in the sample data,k-means algorithm was used to determine the boundary value of off-line driver intention state,and a two-layer hidden markov model was built to determine the parameters of driver intention model.Finally,the off-line simulation verification is carried out by combining the model parameters of intention recognition with some sample data.The simulation results show that the algorithm has an ideal recognition accuracy in driver intention recognition.In order to enable people to participate in the unmanned driving control loop,a path tracking control strategy based on the torque loop control of steering motor is firstly designed.Then,based on the torque loop control,the intervention controller of steering guiding torque is designed by judging the driver's intention to complete the guiding steering torque control of man-machine co-driving.Is adopted in this paper,firstly,consider the nonlinear characteristics of the vehicle tire magic tire model for nonlinear tire lateral force of circumstances,on the basis of the two degrees of freedom vehicle model combined with the steering column model,uses the Receding Horizon control algorithm to derive steering control strategy which based on direct torque control,using the moving serpentine line and double line change two kinds of process to simulate the Receding Horizon algorithm.In order to verify the control effect of the steering torque loop better,the high adhesion condition and the low adhesion condition were verified.At the same time,the PID controller was designed and compared with the Receding Horizon controller.The simulation results show that the Receding Horizon control algorithm has a good effect in terms of stability and tracking accuracy.In order to verify the steering guiding torque control of human-machine co-driving based on driver intention recognition,a Hardware in-loop(HIL)test platform for humanmachine co-driving and completely unmanned driving was established.Firstly,the current loop control of the steering motor,namely the torque loop control,is completed to facilitate the intervention of the driver's hand power,and the control effect of the current loop is verified.In order to achieve pure unmanned driving,accomplishing the steering motor corner ring control,the corner tracking effect is investigated.In order to better simulate the real vehicle environment and the ideal steering resistance moment curve calculated by the Carsim model is tracked by the steering resistance loaded by the double servo motor,the real vehicle steering resistance simulation is realized.In the experiment,path tracking control algorithm is combined with motor control,the torque loop path tracking control system is built,path tracking HIL test using dSPACE/MicroAutoBox and NI/PXI hardware is designed based on the torque loop control.The experimental results show that the path tracking control algorithm can satisfy the requirement of tracking precision,anti-jamming ability and stability.In order to realize the driver's intention recognition,the driver's input angle and rotation speed information are collected by the sensor on the test bed.The HIL test of the driver's intention recognition is designed by using dSPACE/MicroAutoBox and NI/PXI hardware.The experimental results show that the driver's intention algorithm can meet the requirements of the accuracy of intention recognition.The steering torque intervention control experiment based on driver intention recognition was carried out to verify the effectiveness of steering torque intervention control.