Research of Genetic Algorithm in Evolvable Hardware Technology
Lou Jianan,Li Chuantao,Yu Jianhua,Chang Xiaolong
DOI: https://doi.org/10.3969/j.issn.1671-4598.2013.08.074
2013-01-01
Abstract:In addition to the circuit model and chromosome coding's influence on the speed of evolution,the evolution algorithm has a close relationship with speed in evolvable hardware technology.The core idea of the(1+λ)traditional algorithm very suitable for evolution of hardware technology,But through a lot of the evolution of the different target circuit detection,(1+λ)Algorithm has very serious problem about the resistance of the random number seed,Which correspond to different random number seed goals circuit evolution algebra is very different.Based on immune clonal selection principle and the analysis of experimental data,this article improved the algorithm of(1+λ), and through a lot of the experiments proved that the improved method,compared with traditional methods,has significantly improved the speed of evolution,and more concentrated the distribution of evolution algebra.