具有预测功能的规划识别算法研究

A Plan Recognition Algorithm Having Forecast Function

对规划识别中各种事件之间的关系进行了详尽的分析,根据分析结果,给出了具有预测功能的规划识别算法的知识图和规划识别算法,新的规划识别算法采用了自下而上和自上而下相结合生成解图的方法,对Kautz方法、基于规划知识图的方法和具有预测功能的规划识别算法在知识表示结构、搜索方法、冲突解决和求解结果等方面进行了对比和讨论.最后给出了实验结果,结果显示随着领域知识的增加识别算法所需的时间线性增加.

The event relations in plan recognition are studied deliberatively.It presents the plan knowledge graph and recognition algorithm that can forecast future actions.The new algorithm is based on plan knowledge graph and integrates the bottom-up and top-down methods to construct solution graph.The comparisons and discussions of Kautz＇s method,plan knowledge graph and the new algorithm are given on structure of knowledge,search method,conflict solution and result.At last,it gives the experimental results.As the experimental results show,the algorithm is linear-time with the domain knowledge.