本研究之目的在於建構拔河選手個別姿勢之水平拉力關係,以探討整合團隊之拔河表現,利用排列組合所得之團隊最佳值作為以身高、體重、肌力表現等因素進行位置排序之參考依據。受試對象為曾獲得高中組冠軍之團隊七名成員為對象。經由JVC9800U數位攝影機獲取不同拔河姿勢之影像資料,拉力值的獲取是由Biovision多功能生物迴饋系統,並透過visual basic 程式設計軟體所撰寫之數位化視窗軟體TR01將影像資料進行數位化。 經由實際實驗數據採用倒傳遞網路演算法推估受試者於不同拔河姿勢之水平拉力值,結果顯示實際值與推估值之估計標準誤差值不超過5%顯示本研所採用的倒傳遞網路演算法具有極佳的適切性。對於在相關拘束條件下能產生最大拉力值之最佳隊型編排組合,本研究提供依不同隊形可配合成員不同身高、體重及肌力表現進行團隊最佳隊形編排之具體方法。 關鍵詞:拔河、類神經倒傳遞網路演算法
The purpose of this study was to establish a nonlinear estimate model of pulling force in order to estimate the horizontal pulling force by kinematic parameters of individual postures of tug of war. The model was built by the Back-propagation Neural Network. Subjects are seven of the elite high school players whose were 17.5+0.2 years old, 68.2+8.6 kg weight, 173.4+4.6 cm height. The horizontal component force obtain by the load cell of biovision-multi-feedback system. The kinematics data was gained by the TR01 professional digital windows program of visual basic. The standard error of estimate of horizontal force value obtained between the experiment and the model was nor more than 5% of average force. Therefore the model establish by this study could provide a good accuracy. For the related constrains of the optimal team formation that can produce the maximal resultant pulling force, the study provide a complete method to gain the best sequence of players position by different heights, weights, and strength performance. Key words : Tug of war、Back-propagation Neural Network
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。