Joint Learning of Degradation Assessment and RUL Prediction for Aeroengines via Dual-Task Deep LSTM Networks

Health assessment and prognostics are two key tasks within the prognostics and health management frame of equipment. However, existing works are performing these two tasks separately and hierarchically. In this paper, we design and establish dual-task deep long short-term memory networks for joint learning of degradation assessment and remaining useful life prediction of aeroengines. This enables a more robust and accurate assessment and prediction results making for the increment of operational reliability and safety as well as maintenance cost reduction. Meanwhile, the target label functions that match the network training are constructed in an adaptive way according to the health state of an individual aeroengine. Experiments on the popular C-MAPSS lifetime dataset of aeroengines are employed to verify the accuracy and effectiveness. The performance of our proposed work exhibits superiority over other state-of-the-art approaches and demonstrate its application potential.