GM(1,1)模型、Kalman滤波和自适应Kalman滤波在建筑物变形中的应用比较分析

Comparative Analysis of GM(1,1)Model,Kalman Filter and Adaptive Kalman Filter Applied for Building Deformation Monitoring

建立合理有效的模型对建筑物的变形进行分析是至关重要的。通过具体工程实例分析了GM(1,1)模型、Kalman滤波以及自适应Kalman滤波各自的特点:GM(1,1)模型较为适合变形趋势呈指数或线性分布的变形分析与预测,而Kalman滤波对建筑物变形趋势呈线性或波形具有较好的预测效果;在相同的条件下,自适应Kalman滤波较Kalman滤波模型的收敛性及自适应性好,更接近于实际,具有一定的应用效果。

It is critical for building deformation analysis to establish a reasonable and effective model. The characteristics of GM(1,1)model,Kalman filter and adaptive Kalman filter are concluded through analyzing the engineering application. GM(1,1)model is verified to be suitable for the analysis and prediction of deformation with the trend of exponential distribution and linear distribution, while Kalman filter fits for the precision of building deformation with the trend of linear distribution and wave distribution. The adaptive Kalman filter yields better performance than Kalman filter in convergence and adaptability under the same situation,and can be a reference for the application of building deformation monitoring.