论文部分内容阅读
多变量灰色MGM(1,n)模型作为GM(1,1)模型的扩展和补充,弥补了灰色理论未能反应多个变量间的相互影响、协同发展的问题。在遗传进化过程中,存在局部早熟收敛和收敛速度慢的问题。本文利用线性调整法改进遗传算法的适应度值较高的问题,并提出一个对Pc和Pm自适应调整公式,解决了在进行复杂多变量多参数优化问题时,效率低下问题。并利用改进的遗传算法对多变量MGM(1,n)模型的参数进行优化,构建基于改进的遗传算法的MGM(1,n,q)模型。
As an extension and complement to the GM (1,1) model, the multivariable gray MGM (1, n) model makes up for the fact that the gray theory fails to reflect the interaction and cooperation among multiple variables. In the process of genetic evolution, there is the problem of local premature convergence and slow convergence. In this paper, the linear adjustment method is used to improve the genetic algorithm with high fitness value. A formula for adaptive adjustment of Pc and Pm is proposed, which solves the problem of low efficiency in complex multi-variable multi-parameter optimization. The parameters of the multivariable MGM (1, n) model are optimized by the improved genetic algorithm, and the MGM (1, n, q) model based on the improved genetic algorithm is constructed.