Anomaly Detection for Industrial Control System Based on Autoencoder Neural Network

Algorithm 1: Anomaly detection.

Input:

 Input time series ; reconstruct time series ;

 Predict time series ; error check threshold ;

Output:

 Anomaly result ;

1: Calculate sum of error by (2), (3), (4)

2: Calculate smoothed error by EWMA methods (5)

3: Calculate -power error by (8)

4: Mark the anomaly by (9)

5: Return the anomaly observations

Algorithm 1: Anomaly detection.