Quad-RRT: A real-time GPU-based global path planner in large-scale real environments

作者：

Highlights：

• Unlike B-RRT, quad-RRT builds four RRT-trees over the map to speed up the process.

• A complete study of parallelism has been made to analyze the quad-RRT performance.

• Quad-RRT parameters have been tuned up in order to optimize computation time.

• Quad-RRT is 10x faster than the best existing RRT-based approaches.

摘要

•Unlike B-RRT, quad-RRT builds four RRT-trees over the map to speed up the process.•A complete study of parallelism has been made to analyze the quad-RRT performance.•Quad-RRT parameters have been tuned up in order to optimize computation time.•Quad-RRT is 10x faster than the best existing RRT-based approaches.

论文关键词：Global path planning,Large-scale environment,Rapidly-exploring random trees,Graphics processing unit

论文评审过程：Received 19 August 2017, Revised 19 January 2018, Accepted 22 January 2018, Available online 3 February 2018, Version of Record 3 February 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.01.035