Cyberattack Detection in a Robotic Network Using the LPV Kalman Filter Approach

Coordinating a network of robots to achieve consensus is crucial for collaborative tasks. However, identifying faults within such a network is even more critical. This paper investigates a group of Wheeled Mobile Robots (WMR) organized as a master-slave structure, with their interactions described through graph theory. Our goal is to enable the diagnosis of faults within the robots, allowing faulty units to be removed from the formation while ensuring the task continues uninterrupted. To achieve the goal, the LPV Kalman Filter (LPVKF) is utilized to assess the state of every robot, generate residuals to determine if there are faults. The effectiveness of this approach is demonstrated using a Matlab simulator.