基于矩阵编码的遗传算法研究

摘要

This paper analyzes the deficiency of genetic algorithm in solving matrix variable function, proposes a new genetic algorithm based on matrix coding. The selection, crossover, mutation arithmetic operators are defined, and also all arithmetic operators' Matlab function are compiled. It proves the practicability of the new algorithm by simulating a two matrix variables function optimization. Results show that the new genetic algorithm can ensure the configuration integrality of the matrix chromosome, and increase the operation speed and optimization precision.%分析遗传算法求解矩阵函数的局限性,提出一种基于矩阵编码的遗传算法.定义该算法的选择算子、交叉算子、变异算子,编写各算子的Matlab函数,通过仿真求解二矩阵变量函数.实例证明,该算法能确保矩阵染色体的结构完整性,提高遗传算法的速度和寻优精度.