Abstract
Multi-agent systems face challenges in state measurement due to practical limitations and the need for rapid task completion. While traditional finite-time observers address these challenges, they often neglect transient performance and require enhancements in adaptability and robustness, which are crucial for many applications. To address these limitations, this paper introduces a prescribed performance observer that operates without relying on approximation methods, specifically designed for multi-agent systems with uncertainties and disturbances. Simulation results using unmanned aerial vehicles (UAVs) demonstrate that the proposed observer effectively estimates the leader's states and adjacency errors within the system. Compared to traditional finite-time observers, the proposed approach significantly enhances robustness, adaptability, and convergence rate while reducing overshoot for multi-agent systems.