Application of advanced tree search and proximal policy optimization on formula-E race strategy development

作者：

Highlights：

• Targeting the most popular strategic topic in Formula E championship races.

• Proposed two enhancement techniques to Monte Carlo Tree Search.

• MCTS performance remarkably improves using Bivariate Gaussian distribution.

• Integrating MCTS and PPO significantly improves race time solution and consistency.

摘要

•Targeting the most popular strategic topic in Formula E championship races.•Proposed two enhancement techniques to Monte Carlo Tree Search.•MCTS performance remarkably improves using Bivariate Gaussian distribution.•Integrating MCTS and PPO significantly improves race time solution and consistency.

论文关键词：Energy management,Formula-E race strategy,Monte Carlo Tree search,Proximal policy optimization

论文评审过程：Received 14 April 2021, Revised 6 January 2022, Accepted 21 February 2022, Available online 25 February 2022, Version of Record 1 March 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116718