Novel knowledge-based real-time control: Concept, architecture, problem reduction, and representation

This paper explores novel knowledge-based real-time control. First, three control modularizations are originally proposed. Then, by applying problem reduction principle, the knowledge of solving real-time control problems is organized into a hierarchy of reduction from the total problem, which is represented by use of an AND/OR graph. Next, fuzzy linguistics is introduced into the common problem reduction AND/OR graph to consider linguistic fuzzy knowledge of solving real-time control problems, and fuzzy logic is utilized to solve the extended control reduction AND/OR graph. Finally, the novel knowledge-based real-time control is demonstrated by comparative numerical simulations.