Translation Algorithm for Negative Literals in Conformant Planning

摘要

The encoded negative literals in a conformant planning task will result in increasing state spaces. Getting a compact representation of state spaces is one of the most important issues in conformant planning. In this paper, a translation algorithm for negative literals is proposed to reduce the state spaces in a conformant planning task. The relationship between encoded literals is analyzed in detail. Based on the one-of relaxation technique in domain language, the algorithm is used to express the uncertain initial states and action effects in conformant planning. It converts formula one-of into a set of mutually exclusive literals with the relationship of mutual. The experiment study shows the efficiency of the proposed algorithm in pruning the state space in conformant planning tasks.