摘要
为了实现作战模拟训练效能评估自动化,将人工智能的机器学习技术作为解决自动评估的途径,针对主观评定指标提出一种自动评估方法。依据指标数据相关性,替换主观指标并重新构建近似客观评估体系,基于训练集数据而构建的评估算法模型用来为新的评估指标进行效果自动评分。通过实例表明:自动评估模型在保证泛化能力的同时提高了精度和稳定性,在评估模型应用时无需再依靠主观实施评定,为作战模拟训练自动评估提供了可行思路。
In order to realize the automation evaluation of simulation training results,the machine learning technology of artificial intelligence is used as a way to solve the problem of automatic evaluation,and an automatic evaluation method is proposed for subjective evaluation indexes:according to the correlation of indexes,subjective evaluation indicators are replaced by an approximate objective evaluation system,and the evaluation model based on training set data is used to automatically evaluate new evaluation indicators.Examples show that the automatic evaluation model improves the accuracy and stability while guaranteeing the generalization ability,and the method does not need to rely on subjective performance evaluation,which provides a feasible idea for automatic evaluation of military simulation training results.
作者
徐刚
刘金梅
XU Gang;LIU Jin-mei(Battle Support Experiment and Simulation Training Center,Air Force Logistics College,Xuzhou 221000,China)
出处
《火力与指挥控制》
CSCD
北大核心
2020年第5期162-169,共8页
Fire Control & Command Control
关键词
模拟训练
自动评估
机器学习
评估模型
simulated training
automatic evaluation
machine learning
evaluation model