建筑沉降监测数据处理组合模型研究

A Combination Model Research of Building Subsidence Monitoring Data Processing

文章以线性回归分析、对数函数回归分析和灰色系统理论分析这三个子模型为基础,根据最小二乘准则原理,以预测误差为目标函数的误差平方和线形组合构建目标函数,求得加权系数后进行组合预测。通过工程实例说明了组合模型的预测结果优于其他的单项模型预测结果。

The combination model is based on linear regression analysis, logarithmic function regression analysis and grey system theory analysis. Using the principle of least square criterion and the objective function which could be built by linear combination and sum of squared errors aiming to prediction error, the weighting coefficient is obtained. Then building subsidence would be forecast. An engineering example is given to illustrating that combined model predicted results is superior to the other single model predicted results.