Fault diagnosis for hydraulic pump based on intrinsic time-scale decomposition & softmax regression

As the power source of the hydraulic system, the performance of the hydraulic pump is quite crucial for the normal operation of the entire system. The purpose of this study is to propose a method of fault diagnosis for hydraulic pump based on intrinsic time-scale decomposition (ITD) and softmax regression. First, proper rotation components are obtained by processing the pump's vibration data with ITD, which is an efficient and precise time-frequency-energy analysis method of signals. Then the feature vector which contains state information of present moment is formed by the energy values of the components. Finally, the softmax regression model is trained to classify possible fault modes, and the gradient descent method is applied to determine the parameters of the model. The effectiveness and feasibility of the proposed method have been validated by a case study based on the vibration signals of a real hydraulic pump under different fault modes.