Deep Learning Model for Wargame Direct Attack Damage Effect Prediction

In recent years, wargame has become more and more popular across the country, more and more colleges and research institutions have participated in various kinds of wargame competition, our college organized the wargame competition in April, using the MiaoSuan wargame platform developed by the Institute of Automation of the Chinese Academy of Sciences. Through the fierce confrontation in the three stages of warm-up, preliminary and final, the data of 171 wargame replay were obtained. Based on the game replay data, we extract direct attack data to carry out data analysis, data preprocessing, construction and training of damage effect model of direct attack. Through experimental comparison, the superiority of the deep learning model constructed in this paper is verified, which can provide commanders with more accurate prediction in direct attack.