Speed-Scheduled Predictive Yaw Stabilization for a Four-Axle Electric Wheel-Driven Vehicle

Abstract

This paper develops a speed-scheduled predictive yaw-stability strategy for a four-axle electric wheel-driven vehicle subjected to severe steering inputs. A compact lateral–yaw model is updated according to longitudinal speed and used by the vehicle-level controller. Feasible yaw-rate and center-of-gravity (CG) sideslip angle references are generated under a friction constraint, and an incremental model predictive controller computes the corrective yaw moment over a finite prediction horizon. The quadratic program includes bounds on the moment, its rate of change, and the lateral states. An eight-wheel motor-torque allocator then realizes the longitudinal-force and corrective yaw moment demands; under nominal symmetry, it admits a left–right differential-torque solution. PID, sliding-mode control (SMC), and linear-quadratic regulation (LQR) are assessed using the same closed-loop architecture. TruckSim–Simulink simulations with transient steering reversals and sustained sinusoidal steering indicate that MPC yields the smallest peak yaw-rate error. In the sinusoidal case, this error is 0.012 rad/s, whereas PID, SMC, and LQR produce 0.024, 0.018, and 0.014 rad/s, respectively.