Font Size: a A A

The Bayesian Damage Identification Method Based On Additional Virtual Structure

Posted on:2017-05-22Degree:MasterType:Thesis
Country:ChinaCandidate:Z Z WangFull Text:PDF
GTID:2322330488458588Subject:Structural engineering
Abstract/Summary:PDF Full Text Request
In recent years, the large-scale and complex of the civil engineering structures is more and more obvious, so the security and durability of civil engineering structures is payed more and more attention in the very great degree. As an important part of structural health monitoring, structural damage recognition became the focus of research for scholars from all over the world. In the process of structural damage identification, there are always encountering the following problems:(1) It makes damage identification an underdetermined problem that have no enough actual test data; (2) The test can only get the low-order modal information, and it is not sensitive to structural local damage; (3) There are a series of uncertainty factors such as environmental noise, measurement error in the process of test, so it makes the damage identification definite. In order to solve the above problem, this paper put forward a damage identification method based on bayesian theory.At first, this paper gives a brief introduction for the damage identification method based on vibration characteristics and uncertainty damage identification method based on probability statistics, and then analyzes the problems in these methods. Due to the defects in the above method, this paper presents a method for constructing virtual structure based on the virtual deformation method. In this method, it can get new virtual structure through adding virtual structure on the original structure, and then get the low-order modal information which has the higher sensitivity of the structural local damage according to the method of sensitivity analysis. This method can solve the low sensitivity in structural damage identification and the underdetermined problem.In order to solve the uncertainty of damage identification, this paper analyses the probability density function of structural damagement based on the formula of bayesian theory. The structural damage factor can be obtained by the weighted relationship of priori information and test measured modal information. And then use matlab to build up a three layers framework structure to verify the above method. According to this method, the recognition of structural local damage is of great significance. This paper puts forward a method which considers the priori information about local structure itself and others, and then get the probability distribution of the structure damage factor.Finally, according to the above theory, this paper compares the the results of numerical simulation with the practical three layers framework to verify the feasibility of the theory in practical application.
Keywords/Search Tags:Structural Health Monitoring, Damage Identification, Bayesian Theory, Virtual Distortion Method
PDF Full Text Request
Related items