Steering and speed control of a car by fuzzy-neural control

This paper is concerned with driving a car whose control variables are steering and speed. The steering actuator is a stepping motor and the speed actuator is a DC motor. These motors are driven by fuzzy control. The suitable fuzzy control gains for a straight road are different from ones for a curved road. Thus, a neural network is used to judge the grade of a curve in the road. Fuzzy control gains are changed adaptively based on the output of the neural network. The effectiveness of this architecture is demonstrated by experimental tests.