论文部分内容阅读
根据非线性系统状态参数估计理论和土-结构系统的实际地震动记录,利用多项式一阶近似来代替EKF滤波技术中的求导线性化过程,来估计土双自由度集总参数模型中各参数值。这是一种比较新的参数在线实时辨识技术,具有运算简单、计算精度较高的优点。以美国A laska州A nchorage市中心一栋14层办公大楼为算例,结果显示地基土集总参数模型的参数递推结果相当可靠。整个方法易于在M ATLAB操作平台上实现。
According to the non-linear system state parameter estimation theory and the actual ground motion record of the soil-structure system, the polynomial first-order approximation is used instead of the derivation linearization process in the EKF filter technology to estimate the parameters of the soil two-degree-of-freedom lumped parameter model value. This is a relatively new online real-time identification of parameters, with the advantages of simple operation and high accuracy of calculation. Taking a 14-storey office building in downtown Anchorage, Alaska, USA as an example, the results show that the parameter recursion of the aggregate parameter model of foundation soil is fairly reliable. The whole method is easy to implement on M ATLAB platform.