基于力电信号的关节运动意图研究

发布时间：2018-01-11 23:30

本文关键词：基于力电信号的关节运动意图研究　出处：《昆明理工大学》2017年硕士论文　论文类型：学位论文

更多相关文章： 控制肌群 PVDF 压电信号 运动意图

【摘要】：第一次有外骨骼的概念是在1960年左右,但因其涉及的领域太多,所以一直没有太大的进展,而在近十年终于对外骨骼方面有所突破。近年来,外骨骼系统收到了极大的关注,对于能够突破自身极限这点便非常的诱惑人,在众多领域也有着不菲的成就,例如医疗康复训练、身体支撑、运动辅助等方面。人们穿戴着智能外骨骼,一方面能够操保护自己并支撑身体,另一方面也能自由的控制它来完成很多以前一个人做不到的事情。关节是人体中至关重要的结构,本文以肘关节为主要研究对象。肘关节是一个很复杂的结构,在上肢中有着不可忽视的功能,不仅连接手、腕,还与肩关节紧密相连,对人体来说非常重要。肘关节是上肢的中间关节,位置介于上臂与前臂之间,二者构成具有几何形态的机械链,近似一个两脚规样。为了能够完美发挥手的作用,作为主要功能的前臂的旋转运动以及伸臂和屈臂,与肩关节共同协作,保证了手在身体附近的任意位置的自由移动。在人们的日常工作、生活中,肘关节都起着非常重要的作用。例如当人们需要拿取食物,这时肘关节不仅要伸直,还要进行旋前,手便可以拿到食物,然后后屈肘进行旋后的时候,手就可以伸到肩和口的位置,此时食物便能比较方便的被人体摄取;又比如在各种上肢运动的体育赛事中,如若技术动作不够娴熟、训练负荷过载、前期没有足够的热身活动以及肌肉过于疲劳等一系列经常造成肘关节损伤。经研究表明,肘关节周围的控制肌群决定了安全程度。所以研究控制肌群会有相当大的影响,不仅在医学、体育等领域,还能同时预防肘关节的病变,这能起到非常积极的作用。智能外骨骼系统技术的关键点在于识取人体的运动意图,也就是运动信息的获取以及运动意图的识别。采集肌电信号和脑电信号为目前主流的信息获取方式,但这两种采集方式都存在若干缺点,比如易受肌肤导电性、体温变化、心电信号等干扰。另外,由于肌电信号的频谱分布比较窄,而且信号值也很弱,所以在实际采集的时候,信号的一致性和可靠性难以保证。在前期研究的基础上,针对以上提到的问题,进一步建立了以控制肌群的压电信号为输入,关节角度轨迹为输出的肘关节生物力学模型。以单自由度的肘关节屈伸为主要研究对象,搭建针对肘关节的运动信息采集系统,并研究了多种多传感器的数据融合估计算法,再进行仿真比较之后,确定了卡尔曼自适应滤波算法为最合适,最后通过实验进行验证建立的肘关节生物力学模型,将计算得出的关节角度轨迹与实际测得关节轨迹角度相比,发现拟合效果良好,说明通过采集肘关节运动过程中的压电信号可以反映肘关节的运动意图。
[Abstract]:The first exoskeleton concept was around 1960, but it has not made much progress because of its many fields, and has finally made a breakthrough in exoskeleton in recent years. Exoskeleton system has received a great deal of attention, the ability to break through the limits of this point is very tempting, in many areas also have a lot of achievements, such as medical rehabilitation training, physical support. Sports aids. People wear intelligent exoskeletons, on the one hand can operate to protect themselves and support the body. On the other hand, it can be freely controlled to accomplish a lot of things that can not be done by one person before. The joint is the most important structure in human body. In this paper, the elbow joint is the main research object. The elbow joint is a very complex structure. In the upper limb has the function which cannot be ignored, not only connects the hand, the wrist, but also closely connects with the shoulder joint, is very important to the human body. The elbow joint is the upper limb middle joint, the position lies between the upper arm and the forearm. In order to play the role of the hand perfectly, the rotation of the forearm, as the main function, as well as the extension and flexion of the arm, and the shoulder joint work together. The elbow joint plays a very important role in people's daily work and life. For example, when people need to take food, the elbow joint should not only be straightened. Also before the rotation, hands can get food, and then elbow after the rotation, the hand can reach to the shoulder and mouth position, this time food can be more convenient for human consumption; For example, in various upper limb sports events, if the technical movement is not skilled enough, the training load overload. A series of injuries to the elbow, such as insufficient warm-up activity and excessive muscle fatigue, have been reported. The control muscle group around the elbow determines the safety degree. Therefore, the study of the control muscle group will have a considerable impact, not only in medicine, sports and other fields, but also to prevent the elbow disease. This can play a very positive role. The key point of intelligent exoskeleton system technology is to recognize the motion intention of the human body. In other words, the acquisition of motion information and the recognition of motion intention. The acquisition of EMG and EEG signals is the mainstream way of information acquisition, but both of these two methods have some shortcomings, such as vulnerable to skin conductivity. In addition, because the spectrum distribution of EMG signal is relatively narrow, and the signal value is very weak, so in the actual acquisition time. It is difficult to ensure the consistency and reliability of the signal. Based on the previous research, we further establish the control of the muscle group of electrovoltage as the input to the above mentioned problems. The biomechanical model of elbow joint with angle trajectory is output. Taking the single degree of freedom elbow flexion and extension as the main research object, a motion information collection system for the elbow joint is built. Several multi-sensor data fusion estimation algorithms are studied, and the simulation results show that the Kalman adaptive filtering algorithm is the most suitable. At last, the biomechanical model of elbow joint is verified by experiment. Comparing the calculated joint angle trajectory with the actual joint trajectory angle, it is found that the fitting effect is good. The results show that the elbows motion intention can be reflected by collecting the electropressure-electric signals during the elbow motion.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212;R318

【参考文献】