Bidirectional Spectral-Based Transformer for Remaining Useful Life Prediction

The present invention discloses a method for predicting the remaining useful life of an engine based on a bidirectional spectral-based Transformer model (200) comprising: obtaining sensor data from at least one sensor; filtering sensor data; normalizing the sensor data; inputting the normalized data into a bidirectional spectral-based Transformer model (200) by adopting Discrete Cosine Transform (DCT); training the bidirectional spectral-based Transformer model (200) based on a pre-set training data set; and predicting the remaining useful life (108) of the engine based on the trained bidirectional spectral-based Transformer model (200).