Stochastic Model Predictive Obstacle Avoidance with Velocity Reduction in Accordance with Chance Constraints in Crowded Environments

For obstacle avoidance against randomly moving traffic participants, stochastic model predictive control is promising. In crowded environments, however, feasible trajectories satisfying chance constraints do not necessarily exist; crowding induces a relaxation of that causes deterioration safety. To address this issue, we developed velocity method decelerates the ego vehicle to speed satisfies on prediction horizon. We conducted numerical simulations and experiments through crowd comprising vehicles pedestrians evaluate performance. The results indicate designed controller can generate trajectory mitigates adapts various participants.