Wheel Slip Regulation for Heavy Commercial Road Vehicles Using Model Predictive Control Subsumed With Auto-Regressive Time-Series Modelling

Abstract

Wheel Slip Regulation (WSR) is one of the Active Vehicle Safety Systems (AVSSs) for maintaining vehicle stability and maneuverability during emergency braking An approach for wheel slip prediction is proposed in this paper, which involves Auto-Regressive (AR) Time-Series modelling of longitudinal vehicle acceleration. This technique allows the usage of linear longitudinal vehicle dynamics for wheel slip estimation. A wheel slip prediction model was developed considering measurements from accelerometer and wheel speed sensor. This modified the Model Predictive Control (MPC) formulation to a univariate control input problem, involving braking torque. The objective function was devised for solving a least-squares reference tracking problem. An analytical solution for the MPC optimization problem was derived and implemented towards WSR. The proposed framework was programmed in MATLAB Simulink® and co-simulated with IPG TruckMaker® (a vehicle dynamic simulation software). The algorithm was tested in a Hardware-in-Loop (HiL) setup consisting of a pneumatic air brake system interfaced with IPG TruckMaker®. Open loop studies from HiL led to the inclusion of Kalman filter for estimate tuning and PID inner loop control for brake pressure transients, which improved wheel slip regulation.