An Algebraic Estimation-Based Model-Free Control for Acceleration Slip Regulation of Electric Vehicles

This article presents a model-free control (MFC) through ultra-local model and algebraic estimation for a traction control system (TCS). The model-free control uses a first-order ultra-local model, whose parameters are calculated through algebraic estimation and is valid only around the current operation of the system. The basic longitudinal model of the vehicle is presented, emphasizing the wheel slip dynamics and its relationship with the proposed ultra-local model. This relation is used to assess the parameter of the ultra-local model which multiplies the input and is defined as a function of the vehicle speed. The controller performance is demonstrated for different parameter choices and the robustness is verified for subtle changes of the tire-road adhesion coefficient, as the algebraic estimator compensates for uncertainties and deterministic disturbances and leads the system to the desired slip. The controller is compared with a gain-scheduled proportional-integral-derivative (GS-PID) control with gains scheduled in the vehicle speed. Although MFC and GS-PID exhibit similar performance, the primary advantage of the proposed controller is its simplicity and independence from physical models.