Backward-chaining flexible planning

Abstract

Different from all other congeneric research carried out before, this paper pays attention to a kind of planning problem that is more complex than the classical ones under the flexible Graph-plan framework. We present a novel approach for flexible planning based on a two-stage paradigm of graph expansion and solution extraction, which provides a new perspective on the flexible planning problem. In contrast to existing methods, the algorithm adopts backward-chaining strategy to expand the planning graphs, takes into account users' requirement and taste, and finds a solution plan more suitable to the needs. Also, because of the wide application of intelligent planning, our research is very helpful in the development of robotology, natural language understanding, intelligent agents etc. © Springer-Verlag Berlin Heidelberg 2006.