Loading...
3 results
Search Results
Now showing 1 - 3 of 3
- Soft flocks for rendezvous and pursuit missions with distributed MPCPublication . Igreja, José; Lemos, J. M.This paper concerns of MPC (Model Predictive Control) solutions for self-organized or soft flock formations of agents while carrying out an assigned mission. A distributed MPC algorithm is proposed to solve the formulated problem. Pursuit missions by agents formations can also be treated with the exact same algorithm. The obtained solution is completely decentralized in terms of pathfinding and mission completion. The coordination is done only by shared information between the agents, without any additional gathering node or other extra features, like a pre-planning or mission control. Collisions avoidance with obstacles and among agents are solved by introducing coupling constraints in the underlying optimization problem. The algorithm follows the protocol of a Stackelberg strategy game with a finite number of plays seeking an equilibrium outcome. Two examples are presented, the first showing a rendezvous mission and a second one where agents pursuit a moving object, illustrating the wide range of applications for the proposed algorithm.
- Analysis and loop-shaping of holomorphic dynamical PDE systemsPublication . Igreja, José; Lemos, J. M.Holomorphic Dynamical systems is a class of infinite dimensional systems with a infinite number of zeros and no poles. Systems in this class are always BIBO stable with an exact finite settling time for the step response, because they are in feedforward mode, without any inherent feedback physical mechanism. In this paper a precise definition of Holomorphic Dynamical systems is given along with a stability result. Time solutions are also discussed and its control is addressed. This kind of systems are often present in PDE modelling of mass and energy transport phenomena for industrial plant units. Examples and one control application are presented, illustrating the approach described.
- Oracle for guidance with deep neural networks in reusable launch vehicle landingPublication . Igreja, José; Lemos, Joao MOracles are of paramount importance for Deep Neural Networks training. In this paper, an oracle developed for landing reusable launch vehicles is created from a linearizing feedback control law that can perform a prescribed landing trajectory tracking. The oracle is then used to train a Deep Neural Network that can be used as a guidance system for landing maneuvers. Verification is performed by Monte-Carlo.