基于背景值优化的MGM(1,m)自记忆耦合模型及其应用  

Background Value Optimization-based MGM(1,m)Self-memory Coupling Model and Its Application

作　　者：张晓蕾 张军[1] 边锦泽 刘海军[1] ZHANG Xiaolei;ZHANG Jun;BIAN Jinze;LIU Haijun(College of Science,Inner Mongolia Agricultural University,Hohhot 010018,China;Editorial Department of Journal of Inner Mongolia Agricultural University,Hohhot 010018,China)

机构地区：[1]内蒙古农业大学理学院,呼和浩特010018 [2]内蒙古农业大学学报编辑部,呼和浩特010018

出　　处：《内蒙古农业大学学报（自然科学版）》2024年第4期66-75,共10页Journal of Inner Mongolia Agricultural University(Natural Science Edition)

基　　金：国家自然科学基金项目(32160332,32160258);内蒙古农业大学学科交叉基金项目(BR231502)。

摘　　要：MGM(1,m)自记忆耦合模型是一种将传统MGM(1,m)模型和系统自忆性原理相耦合的预测模型,相对于传统MGM(1,m)预测模型,其提高了预测精度。本文在MGM(1,m)自记忆耦合模型的基础上,通过分析传统MGM(1,m)模型背景值计算误差来源,以非齐次指数函数来模拟MGM(1,m)模型的一次累加生成序列,再根据新信息优先原理,改进背景值计算公式,建立起基于背景值优化的MGM(1,m)自记忆耦合模型,并将其应用于基坑变形预测当中,以平均相对误差(ARPE)来进行误差分析和精度检验。研究结果表明,所构建的基于背景值优化MGM(1,m)自记忆耦合模型总的模拟预测精度最高,不仅丰富了灰色预测理论,也为多变量自记忆耦合系统改进方面的研究提供一定参考。The MGM(1,m)self-memory coupled model is a prediction model that combines the traditional MGM(1,m)model with the system self-memory principle.Compared with the traditional MGM(1,m)model,it improves the prediction accuracy.Based on the MGM(1,m)self-memorizing coupled model,this paper analyzed the source of error in calculating the background value of the traditional MGM(1,m)model,and used a non-homogeneous exponential function to simulate the cumulative generation sequence of the MGM(1,m)model.The MGM(1,m)self-memory coupling model based on the background value optimization was established,and it was applied to the prediction of foundation pit deformation,and the error analysis and accuracy test were carried out by using the average relative percentage error(ARPE).The results showed that the MGM(1,m)self-memory coupled model optimized based on the background values had the highest overall simulation prediction accuracy,which not only enriched the grey prediction theory,but also provided a certain reference for the improvement of multi-variable self-memory coupled system.

关 键 词：MGM(1 m)模型 自忆性原理 背景值 预测 

分 类 号：N941.5[自然科学总论—系统科学]