Cyber Attack Detection for a Nonlinear Binary Crude Oil Distillation Column

Abstract

Cyber security for Industrial Control Systems (ICS) is increasingly becoming an area of research as advance network technology continues to evolve providing extended connectivity between cyber world and control system hardware in a plant. In this paper, we present attack detection and isolation technique for sensor attacks on a binary distillation column as an important ICS field which can be targeted by attackers. At first, we present a hybrid model of a DC where the DC has been designed in Aspen Plus Dynamics and the control system has been implemented using Simulink. Then, the mathematical model of various sensor attacks which have been considered for the study are presented. Following that an attack detection and isolation technique based on state estimation using Luenberger observer has been presented. Finally we present the results illustrating effect of sensor attacks on DC performance along with validating the effectiveness of the proposed attack detection and isolation method.