PSA Oxygen Generator Air Control Valve Fault Diagnosis Based on Support Vector Machine
Objective To propose a fault diagnosis method,which can effectively diagnose the fault categories of pressure swing adsorption(PSA)oxygen generator air control valve,and prevent the PSA oxygen generator from not working due to the fault of the air control valve.Methods The time domain and frequency domain analysis methods were used to analyze the pressure signals of the air control valve,and the characteristic values representing typical faults were selected,and the selected characteristic values were trained on the support vector machine(SVM)to obtain an accurate SVM model.The SVM model was used to classify the pressure signals of the air control valve.Results Ten air control valves were selected for non-label verification,and 200 sets of data were collected for each air control valve and put into the SVM model for fault diagnosis.The experimental results were exactly the same as those predicted by SVM model,and the accuracy of experimental verification was 100%.Conclusion A fault diagnosis method for air control valve of oxygen generator is proposed based on the combination of time domain and frequency domain characteristics and SVM model.Through this method,the fault categories of air control valve can be effectively and efficiently diagnosed,and the components in air control valve can be accurately replaced or repaired,which provides ideas and methods for the subsequent fault diagnosis of air control valve.
fault diagnosissignal analysischaracteristic value selectionsupport vector machine
万方数据
维普
