A Hybrid Tactical Plan Recognition Method Base on Planning System

Abstract

Taken first-order predicate logic as formal language, this paper first introduces the knowledge representation form in intelligent planning and plan recognition system. A plan recognition model based on intelligent planning is then set up. For the process of solving problems of planning is to search from initial to final state or from final to initial state, an algorithm for producing the plan sequence is described. The Dempster-Shafer evidential theory, an effective inference method with uncertainty, is combined in the solving process to implement the inference with uncertainty of plan recognition. Finally, a tactical plan recognition example is proposed to explain the process of planning inference with uncertainty, which shows the feasibility of using the presented method for solving the problems of plan recognition.