Abstract
A modified electromagnetism-like mechanism (EM) algorithm is proposed to identify structural model parameters using modal data. EM is a heuristic algorithm, which utilizes an attraction–repulsion mechanism to move the sample points towards the optimal solution. In order to improve the performance of original algorithm, a new local search strategy, new charge and force calculation formulas, new particle movement and updating rules are proposed. The test results of benchmark functions show that the modified EM algorithm has better accuracy and faster convergence rate than the original EM algorithm and the particle swarm optimization (PSO) algorithm. In order to investigate the applicability of this approach in parameter identification of structural models, one numerical truss model and one experimental shear-building model are presented as illustrative examples. The identification results show that this approach can achieve remarkable parameter identification even in the case of large noise contamination and few measurements. The modified EM algorithm can also be used to solve other optimization problems.
Highlights
Modeling and identification of structural systems is a fundamental subject in structural engineering.In the past decades, structural system identification has been widely used in structural health monitoring, damage detection, response prediction, vibration control and other fields [1,2,3,4,5,6,7]
If the data consist of several modal data sets, structural model parameter identification is to solve a set of optimal model parameters, so as to minimize the error between the predicted and measured modal parameters
According to the attraction–repulsion mechanism in electromagnetic theory, electromagnetism-like mechanism (EM) algorithm imagines each solution as a charged particle in space, and the charge of each particle is determined by the value of the objective function to be optimized
Summary
Modeling and identification of structural systems is a fundamental subject in structural engineering. In the identification of structural models based on modal parameters, the natural frequencies and mode shapes of the structure are first obtained through experimental modal analysis, and the difference between the predicted modal parameters and the identified modal parameters is minimized by optimization method to estimate the model parameters. An iterative sensitivity-based FE model updating method was proposed by Teughels and De Roeck, in which the discrepancies in both the frequencies and mode shapes obtained from ambient tests are minimized [14].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
