Translation Algorithm for Negative Literals in Conformant Planning

摘要

The encoded negative literals in a conformantplanning task will result in increasing state spaces. Getting acompact representation of state spaces is one of the mostimportant issues in conformant planning. In this paper, atranslation algorithm for negative literals is proposed toreduce the state spaces in a conformant planning task. Therelationship between encoded literals is analyzed in detail.Based on the one-of relaxation technique in domainlanguage, the algorithm is used to express the uncertaininitial states and action effects in conformant planning. Itconverts formula one-of into a set of mutually exclusiveliterals with the relationship of mutual. The experimentstudy shows the efficiency of the proposed algorithm inpruning the state space in conformant planning tasks.