稳健时序分析在建筑物沉降监测中的应用

以高层建筑物为研究对象，提高预报精度为目的，详细研究了时间序列方法建立预报模型的方法，基于测量数据中存在异常值是难免的，将稳健估计方法引入时间序列建模，提出了基于稳健估计的自回归建模方法，采用某仓库两个沉降监测点连续42期观测数据对该方法进行了验证计算与分析，结果表明数据中存在异常值时，稳健比常规自回归模型的预报精度有较明显的提高。

Application of robust time-series analysis in the building settlement detection

Taking high-rise building as research target and improving prediction accuracy as the purpose,this article specifically the establishing method of forecasting model by using time-series method.Since abnormal value is inevitable in measuring data,robust estimation method is introduced into time-series modeling,and autoregressive modeling method is proposed based on robust estimation.This method is testified and analyzed by adopting the detection data of continuous 42 phases of two settlement detection points...