Nonlinear Optimization For Multi-Agent Motion Planning In A Multi-Obstacle Environment

This work presents a potential field-based control approach for multi-agent dynamical systems in a multi-obstacle environment. A strong focus is put on solving the collision avoidance (either with obstacles or other agents) problem, which is essential for any multi-agent control strategy. In here we discuss potential field constructions which are used for penalizing in the cost the collision avoidance conditions. Comparisons through simulations with mixed-integer formulations are provided. At the end, conclusions and possible extensions are presented.