A novel fault diagnosis for hydraulic pump based on EEMD-LTSA and PNN

The hydraulic pump is the core part of the hydraulic system and impacts the performance of hydraulic directly, thus the diagnosis for hydraulic is crucial. To realize the diagnosis for hydraulic pump, a method utilizing the vibration signal which varies with the performance is proposed. First, ensemble empirical mode decomposition (EEMD) is used to decompose the original signal into finite intrinsic mode functions (IMFs), and then the energy values are extracted to form the feature vector. Second, local tangent space alignment (LTSA), a manifold learning method, is applied in dimension reduction. Third, probabilistic neural network (PNN) is employed as the classifier to recognize the fault pattern. Finally, the effectiveness of the proposed method is validated by the experimental data with different faults