Distributed Model Predictive Flocking with Obstacle Avoidance and Asymmetric Interaction Forces

In most of the existing literature on predictive flocking, the characteristic swarm behavior is formulated in terms of cost functions of quadratic optimization problems with attractive and repulsive interaction forces of equal strength. In this paper, we propose a distributed model predictive flocking framework in which attractive and repulsive interaction forces can be tuned independently by implementing the rules of flocking as softened inequality constraints. The presented framework is able to handle input constraints, obstacle avoidance, and the pursuit of group objectives. The performance of the proposed algorithms is validated in simulation.