The Research of Sneak Circuit Analysis Based on Artificial Neural Network
The reliability and security of aerospace system products is very important, and sneak circuit analysis has become a common analysis method. The research of sneak circuit analysis is an important part of reliability engineering. Sneak circuit analysis is moving into a fully automated, intelligent direction, the automated sneak circuit analysis technology based on learning mechanism like artificial neural network becomes an important part. According to the learning, mapping and predictive ability of artificial neural network, it can be used in sneak circuit. A trained neural network can make a correct prediction of system output under a particular input. Clone selection algorithm is a class of learning algorithms based on e volutionary mechanisms. This paper use clone selection algorithm to train artificial neural network and get predictive matrix. With Comparison between the predictive matrix and design matrix of original system, the sneak circuit problem can be found. This papers work shows that the artificial neural network can effectively solve practical problems in the sneak circuit analysis.
Sneak circuit analysis Artificial neural network Clone selection algorithm.
ZOU TAO MA QISHUANG
School of Automation Science and Electrical Engineering, Beijing University of Aeronautics and Astronautics,Beijing, 100191, P.R. China
国际会议
北京
英文
634-638
2011-06-20（万方平台首次上网日期，不代表论文的发表时间）