Distributed Optimization for Uncertain Nonlinear Interconnected Multiagent Systems With Actuator Faults via Adaptive Command Filtered Backstepping

This article proposed a distributed fault-tolerant optimization algorithm for uncertain multiagent systems (MASs) with unknown interconnections and actuator faults. A hierarchical scheme is developed to decompose the global optimization objective into local fault-tolerant tracking tasks. For each agent, an optimal coordinator is devised to generate the desired trajectory, while an adaptive fault-tolerant controller based on command filtered backstepping is designed to steer the agent's output to this trajectory. This controller can effectively compensate for non-estimable filtering errors, unknown interconnections, and time-varying actuator faults. An integrable auxiliary signal, together with an uncertain bound estimation technique, is introduced to ultimately ensure that all agents' outputs converge to the optimal solution of the total objective function. Moreover, only local information flow is used for each agent, which is challenging due to unknown interconnections. Two case studies are used for illustrating the proposed algorithm.