FuzzyCLIPS and neuro-fuzzy term rewriting for power network control

Abstract

FuzzyCLIPS has been used in several commercial systems to allow for a rule based expert system shell that will also provide fuzzy reasoning capability. We discuss the use of FuzzyCLIPS and neuro-fuzzy term rewriting to define the network model in the distribution management system (DMS) for the Indonesian State Power system (PLN). Although, DMS is currently implemented by a mixed CLIPS/C++ system, the rules defining the behavior of the system are developed using the framework of object-oriented term-rewriting (OOTR) to guarantee the asynchronous characteristics of the program's execution. Another advantage of using OOTR is that we describe both the power propagation in the network and the alarm conditions, which have to be monitored in a uniform fashion. We show how implementing a fuzzy solution will not only provide a more intuitive approach to the propagation problem, but also a more accurate solution.