A prediction model for complex equipment remaining useful life using gated recurrent unit complex networks

Abstract

ABSTRACT Complex equipment has the characteristics of diverse feature types, complex internal structures, and timing information coupling. This paper realizes a complex gated recurrent unit (GRU) network that contains monotonicity-Las Vegas wrapper based feature selection and accelerated GRU based RUL prediction. By eliminating useless data and noise data, the input data volume of the prediction model is reduced, and the efficiency and accuracy of the RUL prediction for complex equipment are effectively improved. The experimental results show our method can predict the RUL of complex equipment more efficiently and increase the prediction accuracy of GRU by 18.3%.