Generator maintenance scheduling using a genetic algorithm with a fuzzy evaluation function

In this paper we consider the problem of generator maintenance scheduling ( GMS) in power systems. A genetic algorithm with a fuzzy evaluation function is proposed in order to overcome some of the limitations of conventional m odelling and solution methods. A test GMS problem is formulated with a reliability objective and flexible and crisp constraints. A rule base and fuzzy sets are formulated for the ob jective and the flexible constraint using experience of solutions to the pr oblem, and used to define a fuzzy evaluation function. This is used in a ge netic algorithm (GA) with an integer representation of the schedule to solv e the test problem. The results obtained from the GA with the fuzzy evaluation function are com pared with those obtained using crisp evaluation functions. The comparison shows that the GA with the fuzzy evaluation function is an effective and pr actical approach for finding good solutions for GMS with flexible constrain ts. (C) 1999 Elsevier Science B.V. All rights reserved.