基于深度学习的乒乓球比赛战术分析系统设计  

Design of Table Tennis Tactical Analysis System Based on Deep Learning

作　　者：曹泽强[1] Cao Zeqiang(Yangen University)

机构地区：[1]仰恩大学

出　　处：《哈尔滨师范大学自然科学学报》2024年第6期60-65,共6页Natural Science Journal of Harbin Normal University

摘　　要：为更加准确的对乒乓球比赛中采用技战术动作进行评价分析,通过深度学习对战术动作扰动值进行采集,其中主要包含“发球”“弧圈”“快攻”“扣杀”“挑打”等动作进行处理,进而得出比赛战术分析的权重值.再根据对乒乓球比赛中的得分情况在不同阶段有针对性布置战术.结果发现:1.发抢阶段中,A、E、G场比赛的弧圈、快攻以及扣杀的动作扰动值大于0.9,挑打动作扰动值小于0.9.在B、D、H场比赛中战术动作扰动值均小于0.8.C、F场比赛中,发球动作扰动值均大于1.2;2.接抢阶段中,A、G场比赛中,动作扰动值均在0.8左右.B、C、F场比赛中,动作扰动值均大于1.0.在D场比赛中战术动作扰动值均大于0.9;H场比赛的战术动作的扰动值均小于0.8,说明在接抢段能压制对手.3.相持阶段中,C、D和E场比赛战术动作扰动值均大于1.0.A、B、F、G 4场比赛中,战术动作扰动值均小于0.9.由此可见,在发抢段的A、C、E、F、G场比赛、接抢段B、C、D、F场比赛以及相持段的C、D和E场比赛中均受到动作扰动的极大影响,导致战术因为受到对手压制.H场比赛的战术动作的扰动影响较小,实现战术压制对手.因此,深度学习可有效的完成比赛战术分析,为未来运动赛场的战术策略提供技术指导.In order to more accurately evaluate and analyze the technical and tactical movements used in table tennis matches,the disturbance values of tactical movements were collected by deep learning,which mainly included"serve","circle","fast attack","smash","pick"and so on.Then the weight value of game tactics analysis is obtained.Then according to the score of table tennis match in different stages of the targeted arrangement of tactics.The results show that:(1)the disturbance value of circle,fast attack and smash in A,E and G matches is more than 0.9,and the disturbance value of pick action is less than 0.9.The disturbance value of tactical action in game B,D and H was all less than 0.8.In C and F matches,the disturbance value of serving action was more than 1.2;(2)In the grabbing stage,the disturbance value of the action in the A and G matches is about 0.8.In game B,C and F,the motion disturbance values were all more than 1.0.The disturbance values of tactical movements were all more than 0.9 in game D;The disturbance value of the tactical action in game H is all less than0.8,indicating that the opponent can be suppressed in the snatch section.(3)In the stalemate stage,the disturbance values of tactical actions in C,D and E matches were all more than 1.0.In A,B,F and G matches,the disturbance values of tactical movements were all less than 0.9.It can be seen that he was greatly affected by motion disturbance in the A,C,E,F and G matches in the first phase,the B,C,D and F matches in the second phase,and the C,D and E matches in the stalemate phase,resulting in the suppression of tactics by opponents.The disturbance of the tactical action in the H game has less influence and realizes the tactical suppression of the opponent.Therefore,deep learning can effectively complete the tactical analysis of the game and provide technical guidance for the tactical strategy of the future sports arena.

关 键 词：战术分析系统 深度学习 乒乓球比赛 战术动作 

分 类 号：TP876[自动化与计算机技术—检测技术与自动化装置]