Consensus Algorithm with Input Constraints and Collision Avoidance for Swarming Sunshade Satellites near the L1 point
Abstract: Swarm Control Theory studies how multi-agent systems interact to solve tasks cooperatively. In this thesis it was studied how graph theory and the Dynamic Average Consensus Algorithm can be used to control swarming sunshade satellites near L1 for global temperature control, and if it would be suitable for the purpose.The Consensus Algorithm was chosen due to the desirable traits, aggregation, pattern formation, and high scalability. To limit the amount of data handled by each agent and handle limitations on communications between agents due to large distances, the problem was modeled with a dynamically changing interaction topology, and to limit the speed of the controller a consensus algorithm with input constraints was considered.The stability of the first part of the controller was proven via Lyapunov stability criterion. The performance of the collision avoidance controller was tested through simulations with random initial states which showed that it can avoid collisions while successfully reaching the desired formation. The final acquired control law for the consensus algorithm consists of two terms: one term that ensures the convergence to the desired location of each agent, and one term that ensures no collisions between two agents take place.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)