Improvement of CUSUM algorithm and its application in intrusion detection for industry control systems
Zhang Yun-gu · Metallurgical Industry Automation · 2014
According to high real-time and availability requirements of industrial control systems,the discriminant function of routine CUSUM anomaly detection algorithm is improved by using the statistical analysis method. Finally,the improved algorithm is applied in the intrusion detection of the hotblast furnace temperature control system in iron and steel enterprises. Results show that the improved CUSUM algorithm can eliminate the random deviation influence and realize the online intrusion detection at a low false alarm rate.