Supervisory intelligent control using a fuzzy logic hierarchy

A generic architecture for supervisory intelligent control based on fuzzy logic is presented. The structure consists of a multi-layer controller which incorporates a rule-based fuzzy logic controller for direct control at the lower level. Higher-level supervisory control is used for supervising the operation of the direct controller. Typical tasks of the supervisory controller are monitoring the process and the controller for determining the operating modes, assignment of initial conditions for the local controller, adaptation schemes, and fault detection and diagnosis. The controller structure is described, followed by its implementation on a real-time process to illustrate the performance and applicability, of the proposed structure.