摘要
针对传统灰色GM(1,1)预测模型在建筑物变形监测预报中的拟合精度较差、预测精度较低和预测时间较短的问题,文中以传统GM(1,1)、线性回归和马尔科夫模型为理论基础,构建了灰线性马尔科夫预测模型,并结合某建筑物变形监测的观测数据,运用新陈代谢的计算模式进行预测。结果表明,灰线性马尔科夫预测模型的拟合精度和预测精度优于单一的灰色GM(1,1)预测模型和线性回归预测模型,灰线性马尔科夫预测模型具有预测精度高、预测时间长和稳定性高的优势。
Traditional gray GM （1 ,1） predicting model has the problems of poor fitting accuracy ,lower prediction accuracy and shorter prediction time in deformation monitoring and forecasting of buildings .In this paper ,a combination of the traditional GM （1 ,1 ） model ,linear regression and Markov model is constructed of grey linear Markov model .Combined with observations data of the deformation monitoring of buildings ,the metabolism computing model is used to predict .T he results show that ：the fitting accuracy and model prediction accuracy of the linear Markov model gray model are better than a single gray GM （1 , 1） forecast model and linear regression forecasting model .Gray linear Markov model has the advantages of high accuracy ,long time and high stability of prediction .
出处
《测绘工程》
CSCD
2016年第10期5-9,16,共6页
Engineering of Surveying and Mapping
基金
国家自然科学基金资助项目(41202245
41272373)
河南理工大学骨干教师资助项目(72105/090)
