Model Predictive Control-Based Path Tracking with Four-Wheel Independent Steering, Driving, and Braking Autonomous Vehicles on Low Friction Road

This paper presents a path-tracking controller for four-wheel independent steering, driving, and braking autonomous vehicles. The proposed controller is designed based on the model predictive control approach to consider the low friction condition of the driving road. The reference state for the model predictive controller is defined by using the desired yaw rate driven from the geometry of the reference path. To compare the effectiveness of the model predictive control, the two different models are used to define the optimization problem. The first controller is based on the point mass model to determine the required yaw moment to generate the lateral motion. The actuator inputs are determined by using the fixed efficiency matrix-based control allocation. The second controller is configured to directly determine the required longitudinal and lateral tire forces to follow the reference state. The determined tire forces are converted to the driving torque and steering angle of each wheel by using the linear tire model. Each controller has a constraint for the tire force and yaw moment to maintain the control effort within the friction circle. The simulation results showed that the path tracking performance on low friction roads has been improved under severe maneuver conditions.