Gear and bearing fault detection under variable operating conditions

The majority of the approaches to the fault detection in rotational machines assume constant and known operating conditions. These assumptions are often violated in practice. Therefore, in this paper we propose a set of features that can be utilized to reveal faults in gearboxes while being robust to fluctuations in operating speed and load. The proposed feature set comprises values of two information cost functions calculated from the coefficients of the wavelet packet transform accompanied by the maximal value of the spectral kurtosis. The fault detection capabilities of the proposed feature set were evaluated on a two–stage gearbox operating under different rotational speeds and loads with different types of mechanical faults.