A Fuzzy Driver on the Handling Track

Abstract

Maneuvering a car around a handling track in minimum time is a challenging task for a driver. The car at high speed is a complex nonlinear Multiple-Input-Multiple-Output dynamic system. A driver must spend many hours learning the skills necessary to control this system proficiently. This complex task is a good test for fuzzy logic control, supporting the premise: humans control complex systems using simple rules. A few fuzzy driving rules are devised to operate on the same inputs and outputs a human driver would use. These rules are encoded in a fuzzy rule base and used to control the system. The resulting fuzzy handling controller is demonstrated in a simulation. A single set of rules is shown to perform well on many different track geometries, illustrating the robustness of the system to a changing environment.