多变量灰色MGM(1,n)模型对建筑物的沉降监测将会产生更好的预测结果,但是灰色MGM(1,n)预测模型存在着自身的缺陷,具有系统的误差.基于此,通过对传统MGM(1,n)模型的残差序列进行修正,建立优化的灰色EMGM(1,n)模型.最后,以某建筑物的沉降实测数据为基础,建立灰色GM(1,1)、MGM(1,n)和EMGM(1,n)模型的预测结果并进行比较,结果表明:灰色EMGM(1,n)模型的预测精度优于灰色GM(1,1)和MGM(1,n)模型,新模型使预测结果更加准确、可靠.%The multivariable grey MGM(1,n) model will produce better prediction results for settlement monitoring of buildings.But there are defects for the grey MGM(1,n) forecasting model, including the systematic errors.Based on the modification of the residual sequence of the traditional model MGM(1,n), the optimized EMGM(1,n) model is proposed in this paper.At last, this paper effectively compares the grey GM(1,1) model, MGM(1,n) model and EMGM(1,n) model, based on practical data of some building.The result shows that the forecasting accuracy of grey EMGM(1,n) model is better than grey GM(1,1) model's and MGM(1,n) model's, and the new model can make the prediction more reliable and accurate, which has practical reference value.
展开▼