Distributed control of multi-agent systems: formation strategies and consensus in networked applications

This thesis focuses on two important problems in the area of Multi-Agent Systems (MAS) and distributed control: the coordination of drone swarms for area coverage and video acquisition, and the consensus-based management of backup batteries in DC islanded grids. Both problems are studied with the aim of developing control strategies that are scalable, adaptive, and suitable for real-world applications, respecting specific constraints. In the first part of the thesis, a leader–follower framework based on Model Predictive Control (MPC) is proposed for the coordination of a swarm of drones. Within this framework, a swarm of drones is adopted to perform area surveillance with mounted downward-facing cameras to capture overlapping videos, as well as maintain a distance-based formation while following a predefined path. The framework ensures collision avoidance while sending the captured videos to the Ground Control Station (GCS) for further processing such as video stitching, object detection, etc. To improve the overall coverage, an online path planning strategy is introduced, allowing the leader to detect uncovered areas and guide the swarm accordingly for rectangle shaped target areas. Then, an improved version of the path planner has been proposed, capable of covering any convex target areas. In addition, the approach takes into account the available network bandwidth by adapting the reference flight altitude to maintain the highest possible video quality. Moreover, an event-triggered approach has been suggested for communication among the drones to improve the efficiency of bandwidth consumption. A grid-based coverage method is also presented as an alternative solution, where the agents form a structured configuration around a central leader. For this case, a nonlinear controller is designed and supported by mathematical analysis to ensure proper system behavior. Sensitivity analyses are carried out to better understand the influence of the design parameters on system performance. In the second part of the thesis, the problem of power sharing among backup batteries in DC islanded grids is addressed. A consensus-based control strategy is developed to coordinate batteries with different capacities and initial states of charge, ensuring a balanced distribution of power. To provide a more realistic representation of the system, a converter transfer function is included in the modeling. The results show that the proposed method can achieve reliable and coordinated operation under practical conditions. Overall, the results of this thesis show that combining the (MPC) control method and the (MAS) framework can effectively address complex coordination problems in real-world applications. Specifically, the proposed methods offer practical solutions for applications in drone-based systems and smart energy networks, while maintaining scalability and robustness. Simulation results through MATLAB, Python, and IsaacSim validate the effectiveness of the proposed methods.