(105f) Cyberattack Detection and Handling Strategies with Online Data-Gathering Capabilities Using Lyapunov-Based Economic Model Predictive Control

Cyber-physical systems (CPS) provide an advanced framework for modern industry and next-generation manufacturing that allows enhanced process monitoring and control through the integration of the physical system with computers and multiple communication channels [1]. Major threats to these CPS’s infrastructures are malicious cyberattacks that can exploit vulnerable communication layers of the system and compromise data reliability. In particular, data gathering devices (e.g., sensors) and final control elements (e.g., actuators) are potential targets of cyberattacks [2]. If data is compromised, feedback control may be lost and model identification procedures cannot be performed online. The impacts of sophisticated cyberattacks on a CPS can be deleterious with consequences that range from profit loss and theft of intellectual property to safety risks associated with industry personnel working on site. This calls for urgent advances in the development of cyberattack-handling control schemes that provide strong safety guarantees for CPS's.

Different types of cyberattacks have been discussed in the literature (e.g., [3][4]) to explore their impacts on CPS platforms. As a step toward cyberattack-resilient control structures, [5] and [6] proposed control formulations based on state estimates for detecting cyberattacks. Important advances in cybersecurity and control structures have also been made in the context of model predictive control (MPC [7]), an optimization and model-based control framework that computes optimal control sequences for a process. In [8], for example, several MPC schemes with economics-based stage cost functions, named economic model predictive controllers (EMPC's [9]), have been explored when only false sensor measurements are considered. In [10], a two-fold control structure has been proposed using neural network-based attack detection capabilities, in which the upper layer is a Lyapunov-based MPC designed to provide safety guarantees after attacks are discovered.

Prior works in our group have explored how EMPC, safety and control-theoretic guarantees, and cyberattack detection concepts interact, and under what conditions a cyberattack policy can be flagged. Specifically, we have investigated and developed control/detection strategies for single attack scenarios in which either sensor measurements or actuators are attacked ([11][12][13][14]). We review the detection strategies proposed in [11] in a new lens as they form a baseline for integrated control/detection frameworks for handling simultaneous sensor and actuator attacks discussed in this talk. The single cyberattack-handling methods pose the question of how to develop a control/detection strategy using LEMPC for handling sensor and actuator cyberattacks occurring at the same time. We use a process example to demonstrate that the case of these simultaneous attacks is particularly stealthy and constitutes a major defining feature of a cyberattack compared to faults, making it necessary to examine this case. Simultaneous attacks are particularly challenging since some of the concepts explored for handling sensor measurement cyberattacks only work if the actuator is not under attack and vice-versa. One of the main contributions of this work, which has been discussed in our recent paper [15], is that it elucidates which combinations of sensor and actuator attack-handling strategies can provide safety in the presence of undetected cyberattacks (even if undetected sensor attacks, actuator attacks, or both are happening simultaneously) and why the other combinations of the discussed strategies cannot achieve this goal. In addition to the ability to detect and handle multiple attack events, inspired by the works in [16] and [17], we show how LEMPC can be utilized for online data collection if certain desired information is needed. If information needs to be gathered online (e.g., for an online model discrimination task), a check must be performed to make sure that the collected data is truly representative of the process and not tampered with by a malicious attack. The integration of cyberattack handling and information collection capabilities may enable a secure data-gathering maneuver that can be performed online while probing for cyberattacks. We elucidate important properties and limitations of the cyberattack-handling methods and data-gathering procedure using LEMPC through simulation studies.

References:

1. Davis, Jim, Thomas Edgar, Robert Graybill, Prakashan Korambath, Brian Schott, Denise Swink, Jianwu Wang, and Jim Wetzel. "Smart manufacturing." Annual Review of Chemical and Biomolecular Engineering 6 (2015): 141-160.
2. Tuptuk, Nilufer, and Stephen Hailes. "Security of smart manufacturing systems." Journal of Manufacturing Systems 47 (2018): 93-106.
3. Setola, Roberto, Luca Faramondi, Ernesto Salzano, and Valerio Cozzani. "An overview of cyber attack to industrial control system." Chemical Engineering Transactions 77 (2019): 907-912.
4. Pasqualetti, Fabio, Florian Dörfler, and Francesco Bullo. "Attack detection and identification in cyber-physical systems." IEEE Transactions on Automatic Control 58.11 (2013): 2715-2729.
5. Ding, Derui, Qing-Long Han, Xiaohua Ge, and Jun Wang. "Secure state estimation and control of cyber-physical systems: A survey." IEEE Transactions on Systems, Man, and Cybernetics: Systems 51.1 (2020): 176-190.
6. Cárdenas, Alvaro A., Saurabh Amin, Zong-Syun Lin, Yu-Lun Huang, Chi-Yen Huang, and Shankar Sastry. "Attacks against process control systems: risk assessment, detection, and response." Proceedings of the 6th ACM Symposium on Information, Computer and Communications Security. 2011.
7. Qin, S. Joe, and Thomas A. Badgwell. "A survey of industrial model predictive control technology." Control Engineering Practice 11.7 (2003): 733-764.
8. Durand, Helen. "A nonlinear systems framework for cyberattack prevention for chemical process control systems." Mathematics 6.9 (2018): 169.
9. Ellis, Matthew, Helen Durand, and Panagiotis D. Christofides. "A tutorial review of economic model predictive control methods." Journal of Process Control 24.8 (2014): 1156-1178.
10. Chen, Scarlett, Zhe Wu, and Panagiotis D. Christofides. "A cyber‐secure control‐detector architecture for nonlinear processes." AIChE Journal 66.5 (2020): e16907.
11. Oyama, Henrique, and Helen Durand. "Integrated cyberattack detection and resilient control strategies using Lyapunov‐based economic model predictive control." AIChE Journal 66.12 (2020): e17084.
12. Rangan, Keshav Kasturi, Henrique Oyama, and Helen Durand. "Integrated cyberattack detection and handling for nonlinear systems with evolving process dynamics under Lyapunov-based economic model predictive control." Chemical Engineering Research and Design 170 (2021): 147-179.
13. Oyama, Henrique, Keshav Kasturi Rangan, and Helen Durand. "Handling of stealthy sensor and actuator cyberattacks on evolving nonlinear process systems." Journal of Advanced Manufacturing and Processing 3.3 (2021): e10099.
14. Durand, Helen, and Matthew Wegener. "Mitigating safety concerns and profit/production losses for chemical process control systems under cyberattacks via design/control methods." Mathematics 8.4 (2020): 499.
15. Oyama, Henrique, Keshav Kasturi Rangan, and Helen Durand. "Lyapunov-Based Economic Model Predictive Control for Detecting and Handling Actuator and Simultaneous Sensor/Actuator Cyberattacks on Process Control Systems." Frontiers in Chemical Engineering: 11.
16. Oyama, Henrique, and Helen Durand. "Lyapunov-Based Economic Model Predictive Control for Online Model Discrimination." Computers & Chemical Engineering (2022): 107769.
17. Oyama, H., Leonard, A. F., Rahman, M., Gjonaj, G., Williamson, M., and Durand, H., "On-line Process Physics Tests via Lyapunov-based Economic Model Predictive Control and Simulation-Based Testing of Image-Based Process Control, " American Control Conference, paper 1187, 2022.