Fuzzified Model Predictive Control Allocation for Integrated Control of Ground Vehicles

Ground vehicles are over-actuated systems, a property that is exploited in hierarchical control architectures in order to attain more cohesive handling and improved road holding. This paper presents a novel global fuzzified model predictive control allocation (MPCA) scheme, incorporating actuator dynamics of the active front steering (AFS), differential braking, and active suspension (AS) subsystems to effectively coordinate amongst the various components in a ground vehicle. The fuzzy logic system is established based on the magnitude of deviation of various key states from their reference values, specifically: the side slip angle, yaw rate, and roll angle. It determines the weights that correspond to each subsystem in the optimization problem, which improves the performance of the control architecture as indicated by various performance metrics in comparison to a typical dynamic control allocation (DCA) scheme. The fuzzified MPCA scheme serves as a mid-level controller that follows upper-level proportional-derivative (PD) controllers that determine the desired corrective yaw and roll moments required to improve the lateral, longitudinal, and roll dynamics. The control architecture is tested in the CarSim
^®
-Simulink
^®
environment in double lane change (DLC) and Fishhook maneuvers, which resulted in better-coordinated handling and stability of both lateral and rollover dynamics.