A hybrid control architecture with fuzzy interface for intelligent control

A novel control architecture for hybrid systems with both continuous variable and discrete-event components is proposed. The architecture includes a conventional controller, a supervisor, and a fuzzy interpreter. The conventional controller is used to control the continuous variable part of the system, while the task of the supervisor is to control the discrete-event part of the system and to make decisions on which control algorithm is to be used by the conventional controller. The fuzzy interpreter links the conventional controller to the supervisor. The proposed approach is suitable for systems which operate at very different operating conditions and need to have proper decision-making mechanisms