Self-learning Hierarchical Fuzzy Logic Controller in Multi-Robot Systems

Abstract

In this paper, hierarchical fuzzy logic control systems and genetic algorithms are amalgamated to provide an integrated knowledge base for intelligent control of mobile robots for collision-avoidance in a common workspace. Genetic algorithms are employed as an adaptive method for learning the fuzzy rules of the control system. A two robot system is considered in the plane, each is to be controlled to a separate target whilst avoiding collision. The hierarchical fuzzy logic control system is made up of two layers to reduce the number of control laws to be learnt by the genetic algorithm. In the first layer, ignoring the possibility of collision, steering angles for the control of each robot to their associated target are determined by a genetic algorithm. In the second layer a genetic algorithm is used to determine adjustments of these controls to avoid collision.