Research on Hydraulic Pump Fault Diagnosis Method Based on SSA-KELM

The stability of hydraulic pumps, as the power source of hydraulic systems, is particularly important. In order to improve the accuracy of hydraulic pump fault diagnosis, a kernel limit learning machine (SSA-KELM) fault diagnosis method based on sparse search algorithm (SSA) was proposed. In this method, the original signal was decomposed into several IMF components by CEEMD, and IMF components with high correlation were selected by the principle of correlation coefficient to complete the signal pretreatment. Secondly, the filtered IMF component was input into the time-frequency index model, and the feature vector was screened and dimensionally reduced through the sensitivity analysis of the time-frequency index to the signal, and the fault identification and classification were input into SSA-KELM. The validation obtained that the recognition accuracy of the method is up to 98%, which is higher than that of the ELM method, verifying that the method has a high recognition effect.