Resilient Consensus Control for Heterogeneous Multiagent Systems via Multiround Attack Detection and Isolation Algorithm

This article is concerned with resilient consensus control for a heterogeneous multiagent system (HMAS) in the presence of malicious attacks on sensors. Most existing strategies are dependent on compensation principle resulting in bounded consensus error. To address this issue, a multiround attack detection and isolation (MR-ADI) algorithm is presented, and, with this algorithm, a resilient isolation-based control strategy is developed to achieve output synchronization. In detail, via output regulator equations, heterogeneous followers are transformed into a virtual layer with same output matrices. With the transformation information from neighbors, a distributed attack monitor is constructed for generating feature signals. With the help of such signals from monitors, a centralized MR-ADI algorithm precisely locates paralyzed followers via a supervisory center, and a distributed MR-ADI algorithm is further proposed only with local and neighbor information. In theory, it is proven that the convergence of output consensus error is ensured. Simulation examples are presented to demonstrate the availability of our theoretical results.