摘要
基于多传感器特征级融合研究了TBM拦截效果评估的问题。首先分析并选取了拦截效果特征信息,构建了基于模糊神经网络的拦截效果融合评估模型;之后从模糊化、模糊推理、反模糊3个方面论述了评估模型的实施过程,进行模糊推理时采用一种改进的BP算法完成了推理规则的实现。最后通过实例仿真验证了改进BP算法的有效性,并验证了评估模型的可靠性及良好的融合性能。
The problem of intercepting effect evaluation of TBM based on muhisensor feature- level fusion was studied. Firstly, the feature of intercepting effect evaluation is analyzed and selected, and a intercepting effect evaluation model based on fuzzy neural network is built. Then evaluating process is discussed in three aspects,fuzzification,fuzzy inference and defuzzification,and a improved BP algorithm is applied in fuzzy inference. Finally,the improved algorithm is verified with availability and the model is proved reliability and fusion performances by simulation.
出处
《火力与指挥控制》
CSCD
北大核心
2014年第7期63-67,72,共6页
Fire Control & Command Control
关键词
拦截效果评估
特征级融合
模糊神经网络
BP算法
重心反模糊
intercepting effect evaluation, feature-level fusion, fuzzy neural network, BP algorithm,center of gravity defuzzification
