Structural Damage Identification Based On Wavelet Energy Entropy

Structural damage identification is a challenging task and plays an essential role in structural health monitoring. Plenty of methods of structural health monitoring have been put forward, but there are still many problems with these method when they are put into practice. Nowadays, many methods are linear methods based on the analysis of structural vibration. An improved wavelet energy entropy method is put forward in this paper, and certain new damage indexes are proposed with the whole consideration of other indexes. This study includes several aspects as follows:Firstly, the wavelet energy entropy method which will be further studied in this article is introduced into the structural damage identification of engineering structure.The parameters of wavelet decomposition will be optimized by the wavelet energy entropy method in the first step, which enables the selection of the optimized layers and wavelet basis of wavelet decomposition. The wavelet coefficients obtained from the optimized wavelet decomposition are used to determine the wavelet energy and wavelet energy entropy, which will be the feature value for the damage identification.The results show that wavelet energy and wavelet energy entropy are able to identify the damage in the structure effectively. On the basis, several damage identification indexes are introduced into this study, such as rate of the average wavelet energy entropy, the difference rate of entropy, reliability index, etc, which have been proved to be quite effective.Secondly, a finite element model is built for the numerical simulation and various kinds of damage cases are set up in this part. Similarly, the acceleration obtained by the wavelet decomposition are used for the calculation of the wavelet energy and wavelet energy entropy. The damage identification indexes are calculated in this part to directly describe the damage conditions. According to the analysis, the results demonstrate that the wavelet energy entropy method and the damage identification indexes are able to identify the structural damage effectively and accurately.Lastly, a bridge model is designed and built in the lab in order to further demonstrate the accuracy of the wavelet energy entropy method. The signal of acceleration extracted by the advanced LMS dynamic test and analysis system will be analyzed by the wavelet energy entropy method as well. The results from the experiment are consistent with that of numerical simulation. As a consequence, the wavelet energy entropy is further demonstrated to be an effective method which are able to identify the damage of structure.