中文摘要：

帕金森病作为一种运动调节功能障碍性疾病，会显著地降低人体中枢神经系统对肢体的控制能力。本文将G—P算法直接应用于人体行走时产生的加速度信号的混沌动力学研究。通过对帕金森病患者和健康人群加速度信号的动力学特征进行比较后发现，相对于健康人，帕金森病患者加速度信号的嵌入维和关联维均显著减小。结果表明，帕金森疾病会使患者运动神经控制系统的混沌程度减弱，复杂性降低；而且可以通过对有限长一维加速度信号的重构捕捉患者运动神经控制系统动力学特征的改变。本研究的结论对于人体运动神经控制系统的生理建模以及帕金森病的临床诊断具有一定的理论和应用价值。

英文摘要：

Parkinson＇ s disease （PD） can significantly reduce the ability of the central nervous system to control human body. In this paper, the G-P algorithm was applied to investigate the chaotic dynamic characteristics of the acceleration signal from human walking. In comparison with healthy subjects, obvious decreases in the embedded dimension （m） and correlation dimension （D2） of acceleration signal from patients with PD were observed. Experimental results demonstrated that PD reduced the complexity and chaotic characteristic of the patients＇ motor nerve system, the change of which could be reconstructed through onedimensional acceleration signals with finite length. The result provides a valuable method for both clinical diagnosis of PD and physiological modeling of human locomotion system.