Along with the coming of industry 4.0 era, industrial internet of things (IIoT) plays a vital role in advanced manufacturing. It can not only connect all equipment and applications in manufacturing processes closely, but also provide oceans of sensor data for real-time work-in-process monitoring. Considering the corresponding abnormalities existing in these sensor data sequences, how to effectively implement temporal anomaly detection is of great significance for smart manufacturing. Therefore, in this paper, we proposed a novel time series anomaly detection method, which can effectively recognize corresponding abnormalities within the given time series sequences by standing on the hierarchical temporal representation. Extensive comparison experiments on the benchmark datasets have been conducted to demonstrate the superiority of our method in term of detection accuracy and efficiency on IIOT-enabled manufacturing.

随着工业4.0时代的到来，工业物联网（IIoT）在先进制造中起着至关重要的作用。它不仅可以紧密地连接制造过程中的所有设备和应用程序，而且还可以提供大量的传感器数据以进行实时在制品监视。考虑到这些传感器数据序列中存在的对应异常，如何有效地实现时间异常检测对于智能制造具有重要意义。因此，在本文中，我们提出了一种新颖的时间序列异常检测方法，该方法可以通过站在分层的时间表示上，有效地识别给定时间序列序列中的相应异常。