Nonlinear linearization controller and genetic algorithm-based fuzzy logic controller for ABS systems and their comparison

Abstract

This paper addresses the design and performance of nonlinear controllers for Anti-lock Braking Systems (ABS). The controllers, based on a nonlinear feedback linearization scheme and fuzzy logic control strategies, are designed to solve the wheel slip-ratio tracking problem of a simplified quartercar ABS model, respectively. Given the desired (optimal) wheel slip-ratio, the feedback linearizing controller cancels all nonlinear dynamics and imposes an appropriate linear behaviour on the wheel slip-ratio such that it will track in a desired way. Using the slip-ratio error and the change-in-error measurements, fuzzy logic controllers have also been developed to resolve the wheel slip tracking problem. In contrast to conventional fuzzy logic control systems, a Genetic Algorithm (GA)-based fuzzy logic controller is proposed to solve the same problem in which the explicit knowledge of parameters for the membership functions is deemed unnecessary. Numerical simulations demonstrate the performance of the proposed control schemes subjected to different road conditions and in the presence of variations in system parameters and a bounded control input. The results indicate that superior performance can be obtained using the GA-based fuzzy logic controller when compared with the feedback linearizing controller and the conventional fuzzy logic controller.