改进二进制编码变异策略研究

由于是一种随机优化方法，标准遗传算法存在着一些不足之处，如局部搜索能力差，寻优精度不高，存在早熟收敛等。为了解决这些问题，提出了一种基于二进制编码基因住的变异策略，对编码串中的各个基因住赋予不同的变异率：在进化初期，赋予个体的高位基因以较大的杂交率，这样可以搜索到更大的解空间，提高算法的全局搜索能力；在进化后期已逼近最优解时，降低高住基因的变异率，减小较优个体被破坏的概率，同时提高低位基因的变异率，增强算法在局部范围的搜索能力。优化实例仿真结果表明，同标准遗传算法相比，改进算法具有寻优精度高，稳定性好，收敛性强等优点。

Binary-encoding Genetic Algorithms with Modified Mutation Operation

Being a stochastic search technique, there are some shortcomings in standard genetic algorithm such as poor local search ability and premature convergence. An effective method of adapting mutation probabilities on each genome of binary-encoding genetic algorithm is proposed to overcome these shortcomings. At the beginning stage of evolution, larger values of mutation probability are assigned to those genomes in higher bit positions, so that larger feasible region is explored. When the region in which the global optimum exists is found, mutation probabilities of genomes in higher bit positions are decreased to prevent better solutions from disruption. Accordingly, mutation probabilities of genomes in lower bit positions are increased to improve the local search ability. Experiments results on five multimodal functions demonstrate that proposed genetic algorithm possesses better optimization ability and gets stuck at a local optimum fewer times than standard genetic algorithm does.