基于周期统计平均的结构动力复合反演研究

研究了输入未确知条件下的参数识别、荷载反演问题。根据激励的周期特性，运用统计平均的思想，构造出一类时域识别反演算法。数值算例表明，该算法的参数识别精度高，且具有很强的噪声适应能力和初参数选择鲁捧性。将本算法与全量补偿算法有机会结合，可更好地解决在旋转动力设备周期激振力作用下的结构参数识别及荷载反演问题。

Composite Inversion of Structural Dynamic System Based on Ideas of Periodic Load Statistical Average

The items of structural parameters identification and load inversion without complete input informations are focused on.A new inversion algorithm is proposed whose procedure is described based on the ideas of periodic load statistical average.The noise free measurements as well as the noise included measurements are considered in the numerical analysis processes to verify the robustness of the algorithm.The numerical results show that the new method has favorable characteristics,such as high accuracy,insensitivity to initial parameter guess and adaptability.Combined with the total compensation method,the proposed algorthm is more efficient in parameters estimation of the structure which is excited by period load of rotating machines.