摘要
本文提出了一种基于神经网络的泛化能力来计算多传感器测量值可信度的方法 ,文中使用了两种类型的神经网络 :CMAC和 BP网络 ;并利用 D- S证据理论将多传感器的多次测量在时间域进行融合 ,以获取准确可靠的融合识别结果。仿真实验表明该方法是可行的 ,能有效地提高系统识别率及鲁棒性。
A kind of method based on the generalization ability of neural network(NN) to calculate the confidence of multi|sensor measurement is presented in this paper. Two kinds of networks, CMAC and BP, are both used. And the D|S evidence theory is used to fuse the multi|sensor information in temporal field in order to get the accurate and reliable results. The simulation shows that the method is practicable and it can effectively enhance the efficiency and robustness of the target identification.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2001年第6期652-655,共4页
Chinese Journal of Scientific Instrument
基金
福建省自然科学基金 (A0 0 1 0 0 0 6)
教育部博士点基金(970 3352 6)资助
