Identificação de falhas em eixos de rotores flexíveis utilizando técnicas de otimização

Abstract

This thesis presents a study about the problem of fault identification in rotating machinery, by using optimization techniques. The basics of this work is to consider that the rotor faults entails strain energy change and, consequently, the vibratory behaviour of the system is also changed. The fault identification strategy used in this contribution is based on the minimization of a functional formed by the difference between the dynamic characteristics of the real system and those obtained from the rotor model, the both entail the cracks. The variables to be optimized are the faults parameters (severity , position and fault length). The dynamic characteristics were calculated through the finite element method. A simple model was used to represent the fault in such a way that it's non-linear behaviour was not taken into account. Pseudo-random optimizatiom methods were chosen in this work, due to their robustness in solving complex inverse problems. Genetics Algorithms were the most important optimization strategy used in this thesis. However, simulated annealing was also successfully used. Simulation results demonstrate that the methodology developed is very effective in identifying parameters of flexible rotors.