遗传优化MGM(1，n，q)模型及在城市用水中的应用  

Multi-variable Grey Model (MGM (1,n,q)) Based on Genetic Algorithm and Its Application in Urban Water Consumption Simulation

作　　者：韩雁[1] 许士国[1] 于常武[1] 

机构地区：[1]大连理工大学土木水利学院

出　　处：《系统仿真学报》2008年第17期4533-4536,共4页Journal of System Simulation

基　　金：“十一五”国家支撑计划项目(2006BAB14B05);国家重点基础研究规划项目(973)(2005CB724202).

摘　　要：城市用水量由于受经济、人口、生活水平等多种因素的影响,具有一定的灰色特征。多变量灰色MGM(1,n)模型作为GM(1,1)模型的扩展和补充,能够反映各变量间相互制约、相互促进的关系。遗传算法具有全局最优性和并行性特点,利用遗传算法对多变量MGM(1,n)模型的参数q进行优化,构建了基于遗传算法的MGM(1,n,q)模型。以1990～2003年大连市城市用水为例,对模型进行了验证,结果表明基于遗传算法的MGM(1,n,q)模型优于MGM(1,n)模型,MGM(1,n)模型要优于GM(1,1)模型。Owing to the influence of economy, population, standard of living and so on, the urban water consumption possesses a certain character of grey. As an expansion and complement, the multi-variable grey model （MGM （1, n）） reveals the relationship of restriction and stimulation. Genetic algorithm possesses the characteristics of full optimal and parallel, Through using the genetic algorithm, the parameter q of MGM （1, n） model was optimized, and a multi-variable grey model （MGM （1, n, q）） model based on genetic algorithm was built. Taking the urban water consumption in Dalian from 1990 to 2003 as case, the model was approved. The results indicate that the multi-variable grey model （MGM （1, n）） based on genetic algorithm is better than MGM （1, n） model, and the MGM （1,n） model is better than MGM （1, 1） model.

关 键 词：灰色系统 MGM(1 n q)模型 遗传算法 城市用水量 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]