Title :
An Efficient Differential Evoluiton Method Used for Evolving Negative-Correlation Circuits
Author :
Zaisheng Huang ; Jingsong He
Author_Institution :
Dept. of Electron. Sci. & Technol., Univ. of Sci. & Technol. of China, Hefei, China
Abstract :
The fault-tolerant design of analog circuit is a very meaning thing. Negative-correlation redundant fault-tolerant design is a new approach. This paper proposes a method which uses single population to evolve negative-correlation circuit. Each chromosome represents only one circuit and evolves one circuit that meets requirements each evolution time. Then use the evolved circuits to guide the subsequent evolution process. According to the information of correlation in population, we change the evolved circuits dynamically. Meanwhile in order to further improve the efficiency of the algorithm, increasing diversity of population, we add variable structure operation and self-adaptive operation of genetic operator. When all the negative-correlation circuits are evolved, the algorithm terminates and outputs these circuits. Analog low-pass filter experiment shows that this method can evolve negative-correlation circuits which meet our requirements successfully. Compared with the conventional methods, it can shorten evolution time and improve the evolution efficiency.
Keywords :
analogue circuits; evolutionary computation; fault tolerance; low-pass filters; analog circuit; chromosome; differential evolution method; evolution efficiency; genetic operator self-adaptive operation; low-pass filter; negative-correlation circuit; negative-correlation redundant fault-tolerant design; Analog circuits; Biological cells; Correlation; Genetics; Registers; Sociology; analog circuit design; differential evolution; low pass filter; self-adaptive;
Conference_Titel :
Computational Intelligence and Design (ISCID), 2014 Seventh International Symposium on
Print_ISBN :
978-1-4799-7004-9
DOI :
10.1109/ISCID.2014.172