Distributed feedback optimization for aggregative tracking in fault-tolerant multi-robot systems

This thesis addresses the challenge of coordinating multi-agent systems in safety-critical environments through a novel distributed control architecture. The proposed framework leverages the principles of Distributed Aggregative Optimization and Distributed Feedback Optimization to enable real-time, resilient coordination among autonomous agents. A key contribution of this work is the development of a fault-tolerant mechanism that ensures system stability and mission continuity even in the presence of multiple agent failures. The effectiveness of the proposed approach is validated through extensive simulations and experimental tests on a heterogeneous multi-drone system tasked with the aerial rescue of a ground vehicle. The results demonstrate the robustness and adaptability of the control architecture, highlighting its potential for real-world applications in Search and Rescue (SAR) missions and other high-stakes scenarios.