New Fault Detection Method for Sliding Bearings Using Empirical Mode Decomposition, Genetic Algorithm and Support Vector Machine

摘要

The failures of the sliding bearings in the marine diesel engines may lead to terrible disaster for the ship operations. It is therefore imperative to detect the faults of the sliding bearings in the early stage. However, the fault detection efficiency is affected by the structure parameters of the support vector machine (SVM). Improper SVM parameters may decrease the fault detection precision. To overcome these problems, a new fault detection approach based on empirical mode decomposition (EMD), improved genetic algorithm (GA) and SVM is proposed in this paper. The EMD can deal with the nonlinear and stochastic characteristics of the vibration data of the sliding bearings. Useful fault features may be extracted by EMD. Then, the improved GA used energy entropy to select individuals to optimize the training procedure of the SVM. The effectiveness of the proposed method has been evaluated with the experimental data. The experiment result shows that the proposed method outperforms the standard GA-SVM method with respect to the detection rate.