基于模糊理论与BP神经网络的导弹质量评估研究  

Missile Quality Assessment Based on Fuzzy Theory and BP Neural Network

作　　者：周璐 顾均元 冯玉光 ZHOU Lu;GU Junyuan;FENG Yuguang(Graduate Team One, Naval Aeronautial and Astronautical University, Yantai 264001, China;310 Teaching and Research Section, Naval Aeronautial and Astronautical University, Yantai 264001, China)

机构地区：[1]海军航空大学研究生一队,山东烟台264001 [2]海军航空大学310教研室,山东烟台264001

出　　处：《兵器装备工程学报》2020年第4期52-57,共6页Journal of Ordnance Equipment Engineering

摘　　要：通过经归一化处理后的导弹各项测试数据,建立合理的模糊化隶属函数模型得出相应的测试项目质量状态模糊集合,结合模糊集合与对应项目的权重值采用模糊综合评判法得出导弹系统的质量状态模糊集合。再通过合理的反模糊化过程,得出导弹各个系统的质量评分值,结合专家评分法和BP神经网络模型训练出导弹总体的质量评分网络,用于由导弹各分系统质量评分值向导弹系统质量评分值的传递,对导弹实际质量评定工作具有指导意义。Taking the missile test data as the quality assessment index,the fuzzy set of the quality state of the corresponding test items was obtained by establishing a reasonable fuzzy membership function model for the normalized test data.Combining with the fuzzy set and the weight value of the corresponding items,the fuzzy set of the quality state of the missile system was obtained by using the fuzzy comprehensive assessment method.Through reasonable anti-fuzzification process,the quality score values of each missile system were obtained.Combining expert scoring method and BP neural network model,the overall quality scoring network of missile was trained,which can be used to get the overall quality scoring value of missile from the quality scoring of each system of missile,and has guiding significance for the actual quality assessment work.

关 键 词：测试数据 隶属函数 模糊综合评判 反模糊化 专家评分 BP神经网络 

分 类 号：TJ760[兵器科学与技术—武器系统与运用工程]