Mobile robot navigation using passivity-based MPC

摘要

This paper presents a novel mobile vehicle navigation algorithm based on the stability analysis of the model predictive control approach. The energy-shaping technique is performed with the navigation function to obtain a new virtual vehicle model that generates candidate feasible trajectories for the motion planner. Stability of the nonlinear model predictive control system is obtained by the passivity concept providing a guaranteed task completion. The proposed approach is adapted for a unicycle mobile vehicle but the work suggests that the passivity-based nonlinear model predictive control concept can be adapted for the navigation purposes for a broad range of mobile vehicle models.