Design of Model Predictive Control with Nonlinear Disturbance Observer for Electric Power Steering System

Considering modern vehicles, electric power steering (EPS) system is one of the most important systems to design the vehicle maneuverability. In this paper, we propose a control method to assist human-steering in order to improve the steering feel by using advanced process control with affinity to autonomous machine-steering. The proposed design is used to manipulate the steering angle so as to follow the reference steering angle while minimizing the quadratic cost function. The proposed control framework consists of three controllers: model predictive control (MPC), nonlinear disturbance observer (NDO), and reference generator. The MPC makes the steering angle follow the reference steering angle generated from the reference angle generator. The NDO estimates steering torque and disturbances from road such as self-aligning torque. The reference generator provides reference steering angle from the assist map and steering torque. We also describe how to adjust the cost function to design the steering feel. The performance of the proposed control system is evaluated and compared to the conventional basic assist control through simulation.