部分输入未知条件下结构动力复合反演的分解算法

在充分利用部分输入已确知而部分输入未知的激励特性的基础上，提出了结构动力复合反演的分解算法，该算法从源头上消除了迭代过程中参数识别与荷载反演的相互影响，降低了问题的计算规模。对于线性参数系统，该算法不经过任何迭代计算即可一次性完成结构参数识别及荷载反演。将其与松弛法结合，可解决非线性参数系统的识别问题，与文献4]的方法比较，其收敛速度有显著提高。

Decomposition algorithm on structural dynamic inversion with incomplete input information

A new time domain identification method called Decomposition Algorithm(DA) is proposed in this paper for solving structural parameter identification with incomplete input information,which is applied to the case when the loads acting on some parts of the structure are known.The method obviates entirely the mutual influence between parameters identification and loads inversion,and decreases the calculaton scale of the problem.Using DA algorithm,the unknown parameters and loads of linear parameter system can be identified directly.When combined with Relaxation Method,DA algorithm is also available to nonlinear parameter system.A numerical example of truss bridge is given for evaluating the validity of the method.The noise free measurements as well as noise included measurements are considered in the numerical analysis processes to verify the robustness of the proposed algorithm,and the results show that DA algorithm has favorable characteristics,such as high efficiency,insensitivity to initial parameter guess and adaptability.