Gear wear detection based on statistic features and heuristic scheme by using data fusion of current and vibration signals

Abstract

[EN] Kinematic chains are ensembles of elements that integrate, among other components, the induc-tion motors, the mechanical couplings, and the loads to give support to the industrial processes that require motion interchange. In this same line, the induction motor justifies its importance because this machine is the core that provides the power and generates the motion for the industrial process. But also, it is possible diagnosing other type of faults that occur in other elements into the kinematic chain reflected as problems in the motor operation. With this purpose, the coupling between the motor and the final load in the chain requires in many situations the use of a gearbox that balances the torque-velocity relationship. Thus, the gear wear in this component is addressed in many works, but the study of gradual wear has not been completely covered yet at different operation frequencies. Therefore, in this work it is proposed a new methodology based on sta-tistical features and the genetic algorithms to find out those features that best can be used for detecting the gradual gear wear of a gearbox by using the signals, measured directly in the motor, from current and vibration sensors at different frequencies. The methodology also makes use of the linear discriminant analysis to generate a bidimensional representation of the system conditions that are fed to a neural network with simple structure for performing the conditions classification. There were tested four uniform gear wear conditions, the healthy state and three gradual condi-tions: 25%, 50%, and 75% wear in the gear teeth. Because of the sampling frequency, the number of sensors, the time for data acquisition, the different operation frequencies analyzed, and the computation of the different statistical features, it is generated a big amount of data that need to be fused and reduced. Therefore, the methodology proposed provides an excellent generalized so-lution for data fusion and for minimizing the computational burden required. The obtained results demonstrate the effectiveness in the fault gradualism detection for the proposed approach.