Cooperative Control achieved by Generic Genetic Fuzzy Logic

Abstract

The paper presents three cooperative control strategies with application to the problem of lake system control. The control system is implemented by a distributed set of communicating agents that use fuzzy logic rules to compute the control decisions. In order to improve the global control performance, each agent sends the control context to its neighbours. The information included in the context and the manner the context exchange is made establish the control strategy: implicit cooperation, coordination and explicit cooperation. Improving the control performance involves more terms in premises of the fuzzy logic control rules, increasing the number of logic rules. Hierarchical fuzzy architectures are used to diminish the number of rules for different methods. The unknown fuzzy logic control rules are obtained offline by the generic genetic approach. The simulations results are given for the three cooperative control strategies and a performance comparison is made.