Research On Structure Health Monitoring And Intelligent Diagnosis Techniques

Research of structural health monitoring (SHM) technology has become a hot topic in civil engineering. So it has important theoretical significance and engineering value to take effective measures to diagnose, evaluate and predict precisely the health condition of buildings in service. For the purpose of SHM and damage diagnosis, the damage feature extraction intelligent diagnosis and trend prediction for engineering structure are studied in this paper.In order to extract damage feature, the methods of damage feature extraction are developed based on Ensemble Empirical Mode Decomposition (EEMD) and wavelet packet transform (WPT). The response signals of the ASCE benchmark structure are processed by using EEMD, the intrinsic mode function (IMF) which contains structural damage information are selected; then the selected IMF is decomposed by orthogonal WPT, and also wavelet package energy (WPE) on decomposition frequency bands are calculated to represent the structure condition. The main results are summarized as:(1) EEMD Methods which use the eliminating characteristics of white noise can avoid the occurrence of modes mixing; (2) For different kinds of damage their WPE distributions are different each other, and for a special damage the distribution of WPE is different at the different detection nodes, which can be used as an ideal target for structural damage characteristics.Due to the problems of the sample shortage in damage diagnosis, an intelligent method is addressed based on EEMD, WPT and support vector machine (SVM). The vibration signal is decomposed using EEMD, so the IMF which contains structural damage information are selected, then the selected IMF is decomposed by orthogonal WPT as extracted features, which are input to a multi-classified SVMs to diagnose structure damage. The method still has good adaptability and classification capability in the case of small samples; and obtains higher diagnostic accuracy by using the radial basis function (RBF) as a kernel function; however, using signals from different detection nodes for the same damage, the recognition correct rate of SVM is different.To aim at fixing the uncertainty caused by only using signals from single detection node in structure damage diagnosis, another diagnosis method is presented by means of multi-sensor feature fusion theory. Through fusing feature extracted from several different detection nodes, it can make different information complementary, and reduce the uncertainty of damage detection information. So precision and reliability of the diagnosis information is much more modified and the diagnosis accuracy was improved.Theoretically, structure damage is progressive. In order to monitor the process efficiently, a feature extraction method of structure progressive damage is studied based on EEMD and Hilbert transform (HT). The response signals are processed by using EEMD, the IMF which contains structural damage information are selected; and then the selected IMF is transformed by using HT and instantaneous frequency(IF) are calculated. A single-degree of freedom structure model and a multi-degree of freedom structure model are used to simulate the progressive process in numerical experiments, and applying this method to practical engineering. It is shown that the IF is obviously changed before and after the structure damage occurrence. The IF can accurately reflect the tendency of structure rigidity change, and represent the developing trend of a structure health condition. So it can be takes as a feature index to monitor the structure progressive condition.In order to solve the problem that is difficult to identify the early damage of structure, a trend prediction model of SVM based on EEMD is proposed, where the EEMD method processing stochastic uncertainty signal and the SVM regression solving the small-sample pattern recognition problem are integrated. The prediction of structural engineering simulation data and vibration data shows that the method can predict the trends of structure conditions accurately and precisely.