Abstract:
This paper addresses the problem of state estimation for linear dynamic cyber-physical systems (CPS) that is resilient against malicious integrity attacks on sensors. A r...Show MoreMetadata
Abstract:
This paper addresses the problem of state estimation for linear dynamic cyber-physical systems (CPS) that is resilient against malicious integrity attacks on sensors. A resilient moving-horizon estimation (MHE) scheme is proposed to correctly estimate the states under sensor attacks by exploiting sensor redundancy, and it is optimal with a guarantee of prior knowledge in the form of both state and disturbance constraints. In this framework, the problem is formulated as a multistage optimal control problem from the perspective of probability theory. Then, it is solved by a special kind of optimization, the bi-level optimization, where the upper-level optimization task responds to the optimal state estimation, while the lower-level optimization task excludes the compromised sensors. Moreover, the strategy to reduce the computational burden is to develop a moving horizon approximation that has been used successfully to develop stabilizing estimation strategy. Numerical simulation is provided to illustrate the performance of the proposed state estimation scheme.
Published in: 2017 11th Asian Control Conference (ASCC)
Date of Conference: 17-20 December 2017
Date Added to IEEE Xplore: 08 February 2018
ISBN Information: