本論文主要研究問題為非線性動態系統的外力識別問題。首先將含噪音訊號的二階微分問題轉換成二階常微分方程(ODE),如此會將原問題改寫成一個回復未知外力反算問題的特例。本論文利用擬時間積分法的技巧將ODE方程式轉換為一拋物線型的偏微分方程(PDE),如此便可以導入廣義格林第二定理,並再利用伴隨崔維茲(Trefftz)測試函數的概念推導出位移和外力的邊界積分關係式。接著利用弱形式方法的概念,先分別計算出弱形式之一階微分運算子和二階微分運算子,最後透過閉合係數展開識別法的假設,將未知外力展開為級數解的形式。透過量測位移資料並以矩陣形式逆轉換的技巧,在不需要任何迭代的情況下,將未知外力之係數以閉合解的形式解出,最後將其疊加回復成所求外力。最終可以將弱形式方法應用在噪音干擾下,回復長時間作用下各種非線性系統未知外力的反算問題當中。
In this thesis, we investigate the external force identification problem in a nonlinear dynamic system. For the recovery of unknown external force in the inverse vibration problems, we consider the second-order derivative of a noisy signal as a second-order linear ordinary differential equation (ODE). Then, it will turn the inverse problem into a special case of the recovery of unknown force in a second-order linear system. After that, we transform the linear ODE of motion into a linear parabolic-type partial differential equation (PDE) by the fictitious time integration method (FTIM), and then use the Green second identity to derive a boundary integral equation in terms of the adjoint Trefftz test functions. Further, we derive a weak-form method to compute the weak-form first-order and second-order differentiators (WFFOD&WFSOD) in terms of series expansion, of which the expansion coefficients can be determined exactly in closed-form without needing of any iterations. Finally, we can recover the external force for nonlinear structures within a large time span and under a large noise.