长跑训练的跑步机故障信号特征提取方法

Feature Extraction Method of Treadmill Fault Signal in Long-Distance Running Training

长跑运动员训练用的跑步机发生故障,会导致使用者受伤,甚至出现生命危险,为此提出长跑训练的跑步机故障信号特征提取方法。分析体育训练跑步机的结构,利用小波包准确分解跑步机信号的高频段信息,分析脉冲信号时间间隔变化和相位变化,提取时域特征得到正常脉冲以及故障脉冲信号间差异。通过小波包分解对应输出信号,重构某尺度内所有频带系数,即可构成全新的时间序列,实现提取故障信号特征。实验证明:所提方法能够有效区分正常信号与故障信号,提取的故障信号脉冲波动非常明显。

The failure of running machine used for long-distance runner training will lead to injury and even life danger of the user.Analyzing the structure of sports training treadmill,using wavelet packet to accurately decompose the high frequency information of treadmill signal,analyze the time interval change and phase change of the pulse signal,and extract the time domain characteristics to obtain the difference between the normal pulse and the fault pulse signal.Through wavelet packet decomposition of the corresponding output signal and reconstruction of all frequency band coefficients within a certain scale,a brand-new time series can be formed and the characteristics of the fault signal can be extracted.Experiments have proved that the proposed method can effectively distinguish between normal signals and fault signals,and the pulse fluctuation of the extracted fault signal is very obvious.