Analysis of Dynamic Test Signal for Bolt Using Empirical Mode Decomposition (EMD)

Abstract

Based on the improved empirical mode decomposition (EMD) algorithm, a signal processing method for dynamic testing of the bolt anchorage system is proposed. As a self-adaptive approach, EMD method can be used to decompose the local characteristics of the no-stationary and nonlinear signals into several fluctuations and trend items with different time scales step by step. Defects information of the bolt anchorage system, which gained from signal separation and diagnosis of stress wave reflection with EMD method, is obtained by getting ride of errors from actual measured signals. The conduction results, with a model test shown, demonstrate that this method is a simple and effective approach with adaptive resolution, and can get more accuracy and flexibility.