Conflict-Based Search (CBS) and its enhancements are among the strongest algorithms for the multi-agent path-finding problem. However, existing variants of CBS do not use any heuristics that estimate future work. In this paper, we introduce different admissible heuristics for CBS by aggregating cardinal conflicts among agents. In our experiments, CBS with these heuristics outperforms previous state-of-the-art CBS variants by up to a factor of five.
|Number of pages||5|
|Journal||Proceedings International Conference on Automated Planning and Scheduling, ICAPS|
|State||Published - 1 Jan 2018|
|Event||28th International Conference on Automated Planning and Scheduling, ICAPS 2018 - Delft, Netherlands|
Duration: 24 Jun 2018 → 29 Jun 2018