Adaptive computer-generated forces for simulator-based training

作者：

Highlights：

• Conducted CGF–CGF and CGF–Human experiments using CAE Inc.’s aircraft simulator.

• Direct comparison of quantitative and qualitative results of both experiments.

• Our adaptive CGF out-performs a doctrine-driven CGF quickly.

• Our adaptive CGF performs better against trainee pilot than against veteran pilot.

• Identified issues to address in subsequent work for more effective adaptive CGF.

摘要

•Conducted CGF–CGF and CGF–Human experiments using CAE Inc.’s aircraft simulator.•Direct comparison of quantitative and qualitative results of both experiments.•Our adaptive CGF out-performs a doctrine-driven CGF quickly.•Our adaptive CGF performs better against trainee pilot than against veteran pilot.•Identified issues to address in subsequent work for more effective adaptive CGF.

论文关键词：Simulator-based training,Reinforcement learning,Self-organizing neural network

论文评审过程：Available online 18 July 2013.

论文官网地址：https://doi.org/10.1016/j.eswa.2013.07.004