A generic algorithm for fuzzy constraint satisfaction problems

Abstract

In this paper, we present a generic algorithm of solving a fuzzy constraint satisfaction problem (CSP) based on the notion of an adaptive level cut: the essence of our algorithm is to apply local propagation for solving different level-cut CSPs. The task mainly involves repair operations to speed up the fuzzified local propagation by reusing as much of the computational effort from the previous level-cut as possible. Besides dealing with constraints in a discrete domain, the algorithm can also be extended to handle constraints in a continuous domain, based on the idea of interval-based reasoning.