基于蚁群算法的多变量MGM(1,N)组合预测模型  

Multi-variable MGM(1,N)Combination Prediction Model Based on Ant Colony Algorithm

作　　者：李树 王丰效 LI Shu;WANG Feng-xiao(School of Mathematics and Statistics,Kashi University,Kashi 844000,China)

机构地区：[1]喀什大学数学与统计学院,新疆喀什844000

出　　处：《数学的实践与认识》2021年第14期41-47,共7页Mathematics in Practice and Theory

基　　金：国家社会科学基金项目(11XTJ001)。

摘　　要：为了提高组合预测的精度,利用多变量灰色MGM(1,N)模型建立了组合预测模型.MGM(1,N)模型的初值一般可以利用第i个累加观测数据作为初值条件.利用传统MGM(1,N)模型,新息MGM(1,N)以及相对误差最小的初值优化模型作为单项预测模型,建立了线性组合预测模型,组合模型权系数利用蚁群算法进行优化.最后利用该组合预测模型进行了实例分析.结果表明,基于蚁群算法的多变量灰色组合预测模型具有较高的拟合精度,同时也说明方法的合理性和可行性.In order to improve the accuracy of combination prediction,a combination prediction model was established by using multi-variable grey MGM(1,N)model.The initial value of MGM(1,N)model can generally use the ith accumulated observation data as the initial value condition.The traditional MGM(1,N)model,the new MGM(1,N)model and the initial value optimization model with the minimum relative error were used as the single prediction model.A linear combination prediction model is established.The weight coefficient of the combined model is optimized by ant colony algorithm.Finally,the combination prediction model is used to analyze an example.The results show that the multi-variable grey combination prediction model based on ant colony algorithm has a high fitting accuracy.It also shows the rationality and feasibility of this method.

关 键 词：MGM(1 N)模型 初值优化 组合预测模型 蚁群算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] N941.5[自动化与计算机技术—控制科学与工程]