顾及模型误差的基坑位移监测数据处理方法

由于基坑地质及观测环境的多样性和复杂性,基坑位移监测数据处理的理论模型与实际模型之间存在着模型误差。若模型误差显著,则所得位移拟合模型的精度较差,将影响后继变形分析的准确性。针对这一问题进行了深入研究,提出了顾及模型误差的最小二乘配置法,并指出在考虑观测值中可能存在粗差的情况下应进一步采用稳健估计法改善基坑监测面的拟合精度。推导了顾及基坑位移模型误差的最小二乘配置法的计算公式,说明了计算步骤。通过对多个实例的计算分析可知,该方法对基坑监测面拟合模型的精度增益能达到35%~48%,是一种比较实用的基坑位移监测数据处理方法。

Data Processing of Pit Displacement Monitoring Considering Model Errors

Due to the variousness and the complexity of the geological conditions and the observation envi- ronments of a pit, it is unavoidable for model errors to exist in the data processing model of pit displacement monitoring. If they are significant, the result will be of poor precision and the successive deformation analysis will be incorrect. The problem is discussed deeply in the paper, and a new method to process the pit displacement monitoring data considering the model errors is put foreword, which is based on the theories of least squares collocation and robust estimation. The relative formulae are then derived in detail. At last, with several examples, it is proved effective in improving the precision of the displacement estimation model, i.e. , the precision improvement rate is at a scale of 35%-48%.