Titlebook: Security of Cyber-Physical Systems: State Estimation and Control; Chengwei Wu,Weiran Yao,Ligang Wu Book 2022 The Editor(s) (if applicable)

跑过

Proactive Secure Control for CPSssystem performance can be steadily deteriorated. To solve the problems, this chapter proposes attack detection, isolation and MTD control schemes for CPSs in a united framework. The physical process is described as a linear time-invariant discrete-time model [61, 73]. False data can be injected into

姑姑在炫耀

Deep Reinforcement Learning Control Approach to Mitigating Attacksarning algorithm is proposed to learn the secure policy for CPSs, based on which deep neural networks are constructed and offline trained. In the inference, trained deep neural networks are deployed to output the secure control signal. The main contributions of this chapter can be summarized as foll

Judicious

作呕

Introduction,nd societal impact and potential [70]. The applications of CPSs can range from the military to the civil critical infrastructure, and more. Antsaklis in [6] has scrutinized the relevant definition, applications, and challenges of CPSs and pointed out that CPSs would transform the way that human interacts with the physical environment.

dendrites

Optimal DoS Attack Scheduling for CPSsnumber of channels allowed to be attacked is no greater than . (. means the total number of channels that can be attacked with limited resources) [75]. To clearly show the interaction between the system designer and the attacker, a signal-to-interference-plus-noise ratio (SINR) based model is introduced to describe the communication channels [83].

Minutes

Zero-Sum Game Based Optimal Secure Controlhe attacker intends to maximize, is defined. Combining the dynamic programming and the game theory, the optimal defense policy and the attack scheme are simultaneously derived to achieve a Nash balance. The main contributions of this chapter can be summarized as follows.

民间传说

CEDE

Learning Tracking Control for CPSs model to describe the physical dynamics. A command model is provided to generate the reference signal. An SINR-based communication model is introduced to describe the interactions between the system designer and the adversary. Following the above setup, the control objective of this chapter is given. Next, we give the details.

exclusice

Secure Estimation for CPSs via Sliding Modeoperty well. Secondly, a sliding mode observer-based secure estimation algorithm is presented to simultaneously reconstruct the systems states, malicious attacks and unknown input signals. Thirdly, the convergence of the proposed secure estimation algorithm is analyzed, and corresponding conditions are derived.

Infirm

Conclusion and Further Worknd reinforcement learning, the problems of securing CPS have been thoroughly discussed, and several effective secure algorithms have been proposed for CPS. Specifically, the conclusions of this monograph can be drawn as follows.