Learnable Data-Oriented Controller for ABS in Brake By Wire Vehicles

A fuzzy controller is suited to control Antilock Brake System (ABS) however time complexity of fuzzy controller is high order. Problem Solution Data Structure (PSDS) has been introduced as a data-oriented model of fuzzy controller to reduce the response time. Locally learning of PSDS controller has been left as an open issue. Fuzzy Learning Mechanism (FLM) has been introduced by using fuzzy inference as learning function, but it needs considerable long process time and steady state errors. In this paper an Adaptive Neuro Fuzzy Inference System (ANFIS) Learning Mechanism (ALM) to learn PSDS is presented. The simulation results show better performance of ALM PSDS comparing to FLM PSDS controller in terms of slip ratio control, stop distance, process time and reduction in steady state errors.