Steering Control for Autonomous Vehicle using Model Predictive Controller

In this paper, an effective solution for the steering control system which is indeed an important part to build the autonomous vehicle is presented. A Model Predictive Controller (MPC) is designed to control the steering angle. Vehicle is controlled by varied counting of the steering angles according to the obstacle's position in the lane. The system has obstacle swerving abilities which is supplemented by a LIDAR detector. The proposed setup analyses the space surrounding the vehicle's lane and tracks moving obstacles. The foremost maneuver is to initiate the motive force to quickly move the vehicle to an adjacent lane, drive past the obstacle and move to the initial lane. Other feature incorporated is the Lane Keeping Assist (LKA) system used to detect when the vehicle deviates from the lane and accordingly adjusts the steering angle mechanically. This performance of the controller is evaluated using MATLAB simulations and is implemented on vehicle prototype.