Abstract

The problem of designing distributed control strategies and developing connectivity metrics for multi-agent systems subject to constrained measurements and external disturbances is studied in this work. The constraint on the field of view (FOV) of the sensing devices used in multi-agent networks is ubiquitous in a wide range of applications. Such constraints have a fundamental impact on the overall performance of the network. The consensus and containment problems for a network of single-integrator agents are investigated, where each agent is assumed to have a sensor with a constrained angular FOV. The flocking problem for a network of double integrators with constrained FOVs is then investigated, where each agent is assumed to be equipped with relative distance and bearing angle sensors, with conic-shaped sensing areas of limited visibility. The angular velocity of the FOV of each agent along with the corresponding control inputs are designed such that the flocking objectives are achieved in a certain neighborhood of the desired configuration. A distributed consensus controller for a network of unicycle agents subject to external disturbances in input channels is also developed for two different cases of disturbances with known linear dynamics and unknown disturbances with known upper bounds. Then, a multi-agent system composed of underwater acoustic sensors is considered, where the network is modeled by a random graph. Different notions for the connectivity assessment of the expected graph of a random network are introduced, and efficient algorithms are developed to evaluate them. Simulations are provided throughout the work to support the theoretical findings.