Abstract:In the process of deformation analysis, it is more reasonable and reliable that multiple monitoring points are applied to modeling and analyzing of grey system in the same deformation monitoring network compared with single monitoring point. But the traditional multivariable grey model (MGM(1,n)) has some short comings. This paper proposes a multivariable grey model based on multivariate total least-squares optimization. And the superiority of this model is proved by some examples. Simultaneously, comparing this model with the traditional MGM(1,n) and analyzing the modeling value and predictive value of these two models. The results show that the modeling and prediction accuracy of this model are higher than MGM(1,n) and this model is more suitable for the actual circumstances when the number of data modeling is over four. This model can be referred to the analysis and forecast of deformation monitoring data in further research.
WU Kaiyan,ZHANG Xianzhou,MA Long et al. A Multivariable Grey Deformation Analysis Model Based on Multivariate Total Least-Squares Optimization[J]. jgg, 2016, 36(8): 682-685.