运动员语音音强特征诊断运动疲劳的探索性研究

The Exploratory Research of Exercise Fatigue Diagnosis Based on Characteristics of Athletes Speech Intensity

提高大运动负荷的训练效率需要准确的运动疲劳状态判断,语音音强特征能够表现运动员的疲劳信息,通过语音信号处理、神经网络建模、灰色关联分析等手段,表明语音音强特征作为运动疲劳判断手段具有可行性。建立了运动疲劳神经网络诊断模型,对语音音强数据进行学习和归纳,确定了语音音强特征与运动疲劳状态之间的模式。模型诊断与定性判断的比较结果表明:对于运动疲劳状态判断,音强特征具有较好的判断等级区分度。建议教练员把直觉判断与音强特征结合起来,可以提高对运动疲劳状态判断的效果。

Improvement of sports training efficiency needs to accurately determine the status of exercise fatigue. The characteristics of speech intensity can display fatigue information for training. By means of the speech signal processing, neural network modeling and gray relational analysis, an indication of feasibility for speech intensity to determine exercise fatigue judgments was shown. Neural network diagnosis model for exercise fatigue was established to conduct learning and induction of intensity data, and the pattern was found for speech intensity and exercise fatigue. The results of the comparison between model diagnosis and coaches intuitive judgment showed that the diagnosis model have good evaluation level discrimination. Recommendation of intuitive judgments combined with intensity characteristic is given, and this can improve the effects of the state of exercise fatigue determination.