Stochastic Modeling of Gearbox via Genetic Algorithm

S. Khanmohammedi, G. Alizadeh, A.R. Giasi, and M.M. Ettefagh (Iran)


Gearbox signal modeling, Bounded random process, McFadden model, Genetic optimization.


Determining the gearbox model has a high importance in validation stage of different fault diagnosis process. Most of previous proposed models were deterministic. Meanwhile the stochastic approach for modeling of signals could improve the model accuracy. In this paper the idea of stochastic modeling of amplitude and phase module is applied for gearbox modeling by using McFadden model. Studying the properties of gearbox amplitude and phase modulation signals, it is seen that the bounded sinusoidal stochastic model is suitable for accurate modeling of amplitude and phase module. The proposed stochastic model contains some parameters to be identified. In this paper the genetic identification algorithm has been used. The good accuracy of proposed model is shown by using the Mont-Carlo simulation of gearbox signal.

Important Links:

Go Back