AUTOMATED FUZZY KNOWLEDGE ACQUISITION WITH CONNECTIONIST ADAPTATION

This paper presents an automated knowledge acquisition architecture fo r the truck docking problem. The architecture consists of a neural net work block, a fuzzy rule generation block and a genetic optimisation b lock. The neural network block is used to quickly and adaptively learn from trials the driving knowledge. The fuzzy rule generation block th en extracts the driving knowledge to form a knowledge rule base. The d riving knowledge rule base is further optimised in the genetic optimis ation block using a genetic algorithm. Computer simulations are presen ted to show the effectiveness of the architecture.