Multi-vehicle Conflict Resolution in Highly Constrained Spaces by Merging Optimal Control and Reinforcement Learning

TL;DR

This paper introduces a hybrid approach combining optimal control and reinforcement learning to resolve conflicts among multiple vehicles in highly constrained environments, ensuring collision-free and feasible maneuvers.

Contribution

It proposes a novel method that integrates RL-learned strategies with optimal control to handle complex multi-vehicle conflict resolution in tight spaces.

Findings

Efficient conflict resolution in confined spaces demonstrated in simulations.

Generated maneuvers are collision-free and kinematically feasible.

The approach effectively explores optimal actions for multi-vehicle coordination.

Abstract

We present a novel method to address the problem of multi-vehicle conflict resolution in highly constrained spaces. An optimal control problem is formulated to incorporate nonlinear, non-holonomic vehicle dynamics and exact collision avoidance constraints. A solution to the problem can be obtained by first learning configuration strategies with reinforcement learning (RL) in a simplified discrete environment, and then using these strategies to shape the constraint space of the original problem. Simulation results show that our method can explore efficient actions to resolve conflicts in confined space and generate dexterous maneuvers that are both collision-free and kinematically feasible.