Research on Life Prediction of Oxygen Concentrator Based on Improved VMD-NAR Neural Network

The airborne oxygen concentrator is an important part of the aircraft life support system. The degraded signal has the characteristics of nonlinearity, non-stationary, and high-noise. In view of the above characteristics in the life prediction including low accuracy, poor traceability, poor authenticity and other problems, an improved VMD-optimized NAR neural network life prediction method is proposed. The oxygen partial pressure signal is decomposed by the VMD method optimized by the gray wolf algorithm to remove the noise in the original signal. Then input the decomposed components into the NAR neural network for prediction. The prediction results of each component are superimposed to obtain the prediction of oxygen partial pressure signal. Comparing the results with methods such as EMD-NAR neural network and BP neural network, the life prediction method has better effect.