Situation Matching in the Nonsequential Planning Environment

This paper is concerned with investigations on new programming tools for planning of solutions for complex tasks. The problem of effective representation and management of knowledge in planning systems with alternative nonsequential reasoning is posed. Then we consider a solution of this problem on example of geometric problem solver by means of Rete-algorithm, which is destined for effective many-situation/many-facts matching. This algorithm has good complexity characteristics and can be used for implementation of situation matcher. The last is a part of the problem solver that is concerned with detecting of situations in knowledge base.