摘要
针对现有人体运动能力评价方法的不足,提出基于人工神经网络的人体100m跑运动能力综合评价方法,并结合30名12～17岁少年男性运动员的各项测试指标及100m跑专项成绩建立人体运动能力的神经网络评价模型。将各项测试指标值经过归一化后作为网络输入,人体100m跑运动能力(由100m跑运动成绩评定)作为网络输出,利用训练好的网络模型进行100m跑成绩预测,得到了较好的预测结果。
Aiming at the insufficiencies of existing evaluation method on kinematic ability, the author put forwards comprehensive evaluation method based on neural network. Combining with test indexes and 100 m running performance of 30 young athletes aged from 12~17, the neural network model of kinematic ability is established. All test indexes are normalized and used as the neural network input, 100 m running ability as neural output. Using the exercising network model of 100 m running to predict performance, we can get best predicting result.
出处
《中国体育科技》
北大核心
2003年第2期1-3,共3页
China Sport Science and Technology
基金
国家自然科学基金资助项目(60171018)
项目名称:虚拟人体运动系统研究。
关键词
100m
短跑
运动能力
人工神经网络
评价
选材
100 m
sprint
kinematic ability
manual neural network
evaluation
talent selection
