A Time-Distributed Spatiotemporal Feature Learning Method for Machine Health Monitoring with Multi-Sensor Time Series

• Published in: Sensors (Basel, Switzerland) Vol. 18; no. 9; p. 2932
• Main Authors: Qiao, Huihui, Wang, Taiyong, Wang, Peng, Qiao, Shibin, Zhang, Lan
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 03.09.2018; MDPI
• Subjects: Deep Learning; Fault diagnosis; Fuzzy logic; Health; Humans; International conferences; machine health monitoring; Methods; Monitoring, Physiologic - methods; multi-sensor time series; Neural networks; Sensors; spatiotemporal feature learning; State of the art; Time Factors; Time series; time-distributed ConvLSTM model; Tool wear; Voice recognition; United States > US; China; deep learning; multi-sensor time series; spatiotemporal feature learning; time-distributed ConvLSTM model; machine health monitoring
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s18092932

Data-driven methods with multi-sensor time series data are the most promising approaches for monitoring machine health. Extracting fault-sensitive features from multi-sensor time series is a daunting task for both traditional data-driven methods and current deep learning models. A novel hybrid end-to-end deep learning framework named Time-distributed ConvLSTM model (TDConvLSTM) is proposed in the paper for machine health monitoring, which works directly on raw multi-sensor time series. In TDConvLSTM, the normalized multi-sensor data is first segmented into a collection of subsequences by a sliding window along the temporal dimension. Time-distributed local feature extractors are simultaneously applied to each subsequence to extract local spatiotemporal features. Then a holistic ConvLSTM layer is designed to extract holistic spatiotemporal features between subsequences. At last, a fully-connected layer and a supervised learning layer are stacked on the top of the model to obtain the target. TDConvLSTM can extract spatiotemporal features on different time scales without any handcrafted feature engineering. The proposed model can achieve better performance in both time series classification tasks and regression prediction tasks than some state-of-the-art models, which has been verified in the gearbox fault diagnosis experiment and the tool wear prediction experiment.