一种基于经验模分解的结构微小损伤检测方法

提出了一种检测结构微小损伤的时域方法。首先采用经验模分解技术分解结构发生微小损伤前后在相同激励下的时域响应信号，得到响应信号的各阶固有模式函数，然后用波形指数来对结构响应信号的各阶固有模式函数进行量化。根据结构损伤前后时域响应固有模式函数的波形指数对比，就可容易地对结构发生的微小损伤进行判断。通过仿真算例发现，结构原始响应信号经过经验模分解后得到的固有模式函数对微小损伤敏感度较好，这种对微损伤的敏感特性可采用各阶固有模式函数的波形指数进行量化。因此该方法是一种时域内检测结构微小损伤的有效方法。

A New Time Domain Approach to Detecting Tiny Damage of Structure Based on Empirical Mode Decomposition

In this paper a time domain approach to detecting tiny damage of structure is proposed. When using the conventional structural damage detection method, especially the frequency domain method, the often used characteristic parameters such as modal frequencies, mode shapes and modal damping ratios, are not sensitive to tiny damage occurring in the structure; and this type of approach is invalid unless the damage accumulates to considerable extent. In contrast to that, in the present work, by considering the features of the Empirical Mode Decomposition (EMD) method which can decompose the response signal of a structure into a series of so called Intrinsic Mode Functions (IMFs), some of these IMFs are much more sensitive to tiny change of the signal than the original signal; thus a structural tiny damage detecting approach is configured. In this approach, the IMFs of the structural response signal are quantified by their shape factors, and the simulation example indicates that the shape factors of IMFs change conspicuously with the introduction of tiny damage to the structure; so the method proposed in this paper is valid for detecting tiny damage of engineering structure.