A Method of Training Six Points Footwork of Badminton Based on Q-learning

• Main Authors: Zi-Yi Huang, 黃子羿
• Other Authors: 黃德成
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/cgi-bin/gs32/gsweb.cgi/login?o=dnclcdr&s=id=%22107NCHU5394062%22.&searchmode=basic

碩士 === 國立中興大學 === 資訊科學與工程學系所 === 107 === In recent years, the research projects on badminton sports has been increased day by day, many experts and scholars have tried to improve badminton players by different training. The training of the six points footwork of badminton will affect the performance of the badminton player''s on-the-spot stability. Therefore, the training is widely listed as the most basic horse-step training for the badminton. However, all kinds of methods with the six points footwork training are specially for badminton player, there are few personal training methods, so the training has limited effectiveness. Therefore, this research propose is to provide more professional training with the artificial intelligence technology. The direction of this thesis will be based on the movement sequence of the badminton six points footwork, selecting specific sequences from all the arrangements, arranging into several combinations, and constructing a simulation environment to allow avatar to test after the training, and judge the effect of the training. It will use the Q-Learning algorithm to learn, just like creating a virtual coach, it can help you plan how to adjust the training, and hope to find a suitable training group. It is found that the Q-Learning simulation training method can achieve this goal and make the test scores ideal from the results of the experiment. After many iterations, Q-Learning can learn how to adjust the training faster and continue to find other suitable training combinations.