基于加权模糊Petri网的无人机自主任务推理决策研究  

Reasoning and Decision Making for Autonomous UAVs Based on Weighted Fuzzy Petri Nets

作　　者：谭雁英[1] 童明[2] 张艳宁[3] 祝小平[4] Tan Yanying Tong Ming Zhang Yanning Zhu Xiaoping(Institute of Precision Guidance and Control, Northwestern Polytechnic University, Xi＇an 710072, China The Chinese People＇s Liberation Army 68128 Forces, Lanzhou 730000 School of Computer, Northwestern Polytechnic University, Xi＇an 710129, China UAV Research Institute, Northwestern Polytechnic University, Xi＇an 710065, China)

机构地区：[1]西北工业大学航天学院精确制导与控制研究所 [2]中国人民解放军68128部队 [3]西北工业大学计算机学院 [4]西北工业大学无人机所

出　　处：《西北工业大学学报》2016年第6期951-956,共6页Journal of Northwestern Polytechnical University

摘　　要：围绕无人机搜索/攻击自主任务推理决策典型想定,针对无人机自主任务推理决策规则中存在的不同输入命题对结论的贡献和影响不同,对加权模糊Petri网(weighted fuzzy petri nets,WFPN)形式化推理算法进行了深入研究;设计了双机自主搜索/攻击任务推理决策规则集;为进一步适应上述想定下规则集中存在的多条规则的输出命题是相同的情形,对WFPN形式化推理算法进行了改进;应用分层的策略,采用改进的WFPN形式化推理算法和上述设计的规则集,仿真验证了双机搜索/攻击自主任务推理结果的合理性,并降低了推算过程中矩阵的维数。This paper is dealt with autonomous reasoning and decision-making for unmanned aerial vehicles（ UAVs） executing the search and attack mission. There exists the fact in the above reasoning envision that each input proposition does different contribution to the conclusion of the rule,and so the formalized reasoning method based on weighted fuzzy Petri nets（ WFPN） is studied. The weighted fuzzy production rule set about two UAVs cooperating in search and attack mission is designed. The formalized reasoning method based on WFPN is further improved to adapt to the situation that the output propositions of several rules are the same in the above rule set. Applied the layered strategy,the improved formalized reasoning method is simulated based on the designed rule set. The simulation shows that the reasoning results are reasonable and the dimension of the large reasoning matrix is reduced efficiently.

关 键 词：无人机 自主 加权模糊PETRI网 形式化推理 分层策略 决策 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]