An adaptation method for fuzzy logic controllers in lateral vehicle guidance

A model reference adaptive fuzzy logic controller (MRAFLC) is designed and simulated for a full-sized test vehicle to achieve control of the lateral motion of the vehicle. The structure of the FLC is modularized as a feedback and preview rule base. The parameters of the FLC are tuned automatically using the MRAFLC adaptive tuning scheme. The goal is to make the output of the closed-loop system, under fuzzy logic control, follow a reference output, generated by a fuzzy system. The first step is to define the FLC in a form that enables it to be adjusted on-line. Using Lyapunov theory a supervisory control term is constructed such that the closed-loop system under fuzzy logic control will maintain stability in the sense that the states of the vehicle are bounded by a maximum value prescribed by the control designer. An additional Lyapunov analysis is performed to place sufficient conditions on the stability of the adaptation law, ensuring that the output error term converges to a lower bound.