Research On Guide Steering System With Driving Intention Recognition

With the rapid development of automobile technology,driver assistance systems such as lane keeping play an increasingly important role in the active safety of vehicles.However,because the technology is far from 100% mature,and the auxiliary system is working independently,it is difficult to switch the driving assistance function very smoothly.Therefore,the human-machine driving is a better solution nowadays,and it is a transitional phase in the middle of a complete driverless-driving process.The guided steering system studied in this paper is based on EPS motor,change the motor controller and enables it to perform angle control and torque control to achieve torque guidance.When the driver’s hand is not on the steering wheel,the controller executes angle control to achieve the function of lane keeping.In this case,when the system receives the lane change instruction,it will change the path along the planned path which is obtained by 5-order polynomial,and guarantee to change the lane within a certain period of time.When the driver starts driving,the controller changes to torque control which is exactly guide-steering.In the process of man-machine co-driving,the system adjusts the magnitude of the torque encountered according to the deviation between the driving trajectory and the desired trajectory.When the deviation is within a certain range,the driver will feel very little guiding torque;As the deviation increases,the guiding moment will increase linearly,and the driver will feel the increasing guiding moment;Finally,the guiding torque will be saturated,in which case the driver still has the right to decide the steering wheel angle.In addition,according to the vehicle status,road information and driver’s information,the system will identify the driver’s intention to maintain the current lane or change lanes.After identifying the driving intention,the system will conduct the corresponding lane maintenance or change lane guidance.In order to realize the function of lane maintenance and lane change assistance,the Mobileye on the real vehicle is verified,and the result of verification is compensated by the sensor using the inertial measurement unit,so that the guide-steering system can get stable and continuous road information.To verify the effect of the control algorithm,performance indicators are proposed and compared with the parameters of the driver.Both lane-keeping and lane-changing assistance angle control are validated on the real vehicle.The guiding moment used is related to the difference between the desired angle and the actual angle.The guidance test is carried out on the test bench.In order to realize the recognition of driving intention,the identification model is established by using logistic regression algorithm and the model is trained by collecting data from the test bench.The observed values of the recognition are the difference between the actual and expected corners,the variation rate of the deviation and the average value in a period of time.The result of the recognition is the intention of the driver.Finally,the intention recognition model is verified by real vehicle data.