Secure state estimation for nonlinear power systems under cyber attacks

Abstract

This paper focuses on securely estimating the state of a nonlinear dynamical system from a set of corrupted measurements. In particular, we consider two broad classes of nonlinear systems, and propose a technique which enables us to perform secure state estimation for such nonlinear systems. We then provide guarantees on the achievable state estimation error against arbitrary corruptions, and analytically characterize the number of errors that can be perfectly corrected by a decoder. To illustrate how the proposed nonlinear estimation approach can be applied to practical systems, we focus on secure estimation for the wide area control of an interconnected power system under cyber-physical attacks and communication failures, and propose a secure estimator for the power system. Finally, we numerically show that the proposed secure estimation algorithm enables us to reconstruct the attack signals accurately. which use the synchrophasor technology to maintain the system's stability. WACS employ advanced data acquisition, communications, and control to enable increased efficiency and reliability of power delivery [12]- [14]. While WACS are the most promising technology to detect small signal instabilities in large power systems, their performance is highly dependent on the received data and on the underlying communication network. Clearly, the greater dependence on communication systems increases opportunities for cyberphysical attacks and disturbances. Extensive work has been done on monitoring and on autonomous feedback control for WACS [15], but this work has not studied how to identify cyber-physical attacks or communication failures, and how to perform secure state estimation for WACS. In this paper, we focus on secure estimation for the wide area control of an interconnected power system, and assume that the system operator has installed several phasor arXiv:1603.06894v1 [cs.SY]

Detection of Sensor Attack and Resilient State Estimation for Uniformly Observable Nonlinear Systems having Redundant Sensors

“…However, assuming the measurement output to be a "flat" output limits the class of systems; for example, the given system should not have non-trivial zero dynamics [15]. On the other hand, a secure state estimator was constructed in [16] for a special form of nonlinear systems whose stacked outputs can be represented by a linear function of the initial state and the attack vector.…”

Optimal Node Attack on Causality Analysis in Cyber-Physical Systems: A Data-Driven Approach

“…Thus, the causality of nodes (sensor groups) can be analyzed with the lower calculation burden instead of the CA of measurement data directly from sensors. Therefore, we can evaluate the significance for each node instead of the significance for each measurement variable through ( 9), ( 10) and (11). The validity of this method will be verified in the subsequent experiments.…”

“…Recently, due to the frequent industrial security contingencies in cyber-physical systems (CPS) (such as smart grids and other industrial control systems [1], [2]), the security problems induced by malicious cyber attacks in CPS have received wide attention [2]- [6]. The main research on these security problems can be divided into four categories: attack strategies [1], [7], self-protection of systems [8], attack detection and identification [2], [9], [10], and design of resilience state estimators and controllers [3]- [5], [11]. Note that the last three categories represent countermeasures against different attack strategies.…”

Resilient Cyber-Physical Energy Systems Using Prior Information Based on Gaussian Process

“…In the literature, resilient state estimation has attracted significant attention [10]. While there are numerous works on the topic, most of them focus on attack-resilient algorithms that consider measurement noise [11], [12], time varying attack support [13], robustness considerations [14] and distributed case [15]. There are also numerous applications including power systems [16], UAVs [17], energy delivery systems [18], autonomous vehicles and networked systems.…”

“…We provide theoretical guarantees of how certain boundedness properties of the prior information set can improve the reconstruction error bound of the resulting resilient estimator. Unlike previous work [13], [23], [24] which depends on the restricted isometry property (RIP) [25], we developed our results by leveraging the latent information contained in the auxiliary model. Moreover, the developed estimator is applied on realistic data acquired from the New York Independent System Operator (NYISO) transmission grid.…”