论文部分内容阅读
目的利用改进的多元多尺度熵特征对人体静态平衡运动力学信号进行分析。方法在多元多尺度熵计算中,要对多元延迟向量的维数进行扩展,本文针对性地将单个变量的依次嵌入改成对所有变量同时嵌入。结果改进算法的多元多尺度熵特征应用于多种平衡模式的实验中,处理速度更快,熵值在模式间的距离更大,模式内的散度更小,更易区分。结论本文算法提高了计算效率,改善了特征的可区分性,可以更好地分析人体的静态平衡能力。
OBJECTIVE: To use the improved multivariate multiscale entropy feature to analyze the human body’s static balance kinematic signal. In the multivariate multiscale entropy calculation, the dimension of multivariate delay vector is extended. In this paper, we embed the single variables in turn into embedding all the variables simultaneously. Results The multivariate multiscale entropy feature of the improved algorithm is applied in many experiments of equilibrium mode. The processing speed is faster, the entropy value is larger in the modes, and the divergence in the model is smaller and easier to distinguish. Conclusion The proposed algorithm improves computational efficiency and improves the distinguishability of features, which can better analyze the human body’s static balance ability.