Neural network implementation of a fuzzy logic controller

Fuzzy logic is an attractive technique for plant control but suffers from complex data processing. A solution to this problem is here presented and consists in implementing a fuzzy logic controller (FLC) on a neural network (NN). As an example, a DC drive is considered. After designing a FLC for controlling the drive speed, a NN is trained by supervision to learn the input-output relationship of the FLC. It is found that a NN with few neurons implements the relationship very well. Equipment is then set up to test the NN control. The experimental responses obtained from the drive demonstrate the effectiveness of the solution.<
>