A Novel Negative-Correlation Redundancy Evolutionary Framework Based on Stochastic Ranking for Fault-Tolerant Design of Analog Circuit

摘要：

　　The fault-tolerant evolutionary design based on negative-correlation redundancy technique is an effective way to improve the fault-tolerance of analog circuits with uncertain faults.In the existing negative-correlation redundancy evolutionary framework (ENCF), the negative-correlation penalty coefficient plays an important role, and it affects the performance of ENCF greatly.However, the value of the negative-correlation penalty coefficient is heavily dependent on the experience of designers.In this paper, we propose a new negative-correlation redundancy evolutionary framework based on stochastic ranking strategy.In order to make comparisons with the existing researches, we employ analog filter as a design example.Experimental results show that the framework proposed in this paper can generate negatively correlated redundancies without specifying the penalty coefficient, and it shows a relatively high ability to convergence compared to ENCF.

关键词： Negative-correlation analog circuit fault-tolerant stochastic ranking genetic algorithm

作者: Chao Lin Jingsong He

作者单位: Department of Electronic Science and Technology, University of Science and Technology of China, Hefei, China

会议类型: 国际会议

会议名称: 4th international Conference,ICSI2013(第4届群体智能国际会议)

会议地点: 哈尔滨

会议语种:英文

页码: 556-563

在线出版日期: 2013-06-12（万方平台首次上网日期，不代表论文的发表时间）