康复用下肢外骨骼系统仿生步态规划方法研究

发布时间：2018-02-20 19:19

本文关键词： 康复外骨骼机器人模糊变负载控制步态四足拄拐　出处：《中国科学院深圳先进技术研究院》2016年硕士论文　论文类型：学位论文

【摘要】：随着我们步入21世纪,人口老龄化比例大幅增加,由于偏瘫、中风导致的下肢不便老人不断增多,由于脊柱神经受损而导致下肢无力、偏瘫、瘫痪的人群也在不断增加。自20世纪以来,以美国为首的国家开始不断及进行助力外骨骼以及康复外骨骼机器人的研究,伴随着机器人技术的不断进步,各个国家在下肢康复外骨骼的研究有了不断的突破。我们国家也逐渐认识到机器人对科技实力提升的重要性,逐步开始大力推广和发展机器人事业,近年来不断有新的机器人出现。康复用下肢外骨骼机器人能够帮助患者进行腿部康复训练,训练在可重复性方面优于传统方法,可以大幅缩短患者的康复训练时间,并减少陪同康复医师的数量,大幅降低人力成本。对社会和我国康复医疗,以及机器人发展起着积极的贡献作用,给老年人、下肢不便患者行走提供有效帮助,提高社会弱势群体的生活自主能力,改善他们生活质量。本论文着重研究康复外骨骼仿人步态,是由于该步态不同于自然人的步态,也称这种步态为非常态步态。首先从步态安全性方面进行研究,通过传感器对外骨骼运行状态进行判断,并对步行中发生的意外进行实时判断,防止对穿戴者造成二次损伤。本文所用外骨骼机器人步态使用预设步态轨迹,需要伺服驱动系统能够跟随控制计算机内预设步态进行轨迹。为了改善驱动系统对变负载情况下的动态响应特性,驱动关节使用了关节负载自适应控制。在变负载情况下,这种控制方式比传统PID控制方式有着更好的动态响应特性,可以有效提高伺服系统、机械结构的使用寿命,并提高伺服系统对预设步态轨迹的跟随能力。通过对正常人步态特点的分析和研究,本文结合外骨骼四足拄拐步态的特点,对四足步态进行采集。通过大量实验最可以终得到较为理想的步态轨迹,目前让两位截瘫患者自主行走。本作者在实验期间进行了大量的非常态步态实验,发现穿戴者使用拐杖的方式直接影响步态的稳定性,特别是刚开始学习四足步态的时候,随后提出了四足拄拐长步态。通过让多位实验者进行步态实验,验证了在拐杖支撑不合理的情况下,动态步长步态比传统固定步长步态有着更高的稳定性,并且重心摆动较小,更为省力。
[Abstract]:As we step into 21th century, the proportion of population aging increases dramatically, and because of hemiplegia, stroke causes an increasing number of inconveniences in the lower extremities, and because of spinal nerve damage, it leads to lower extremity weakness, hemiplegia. The number of paralyzed people is also growing. Since 20th century, countries led by the United States have continued and carried out research on assisted exoskeletons and rehabilitated exoskeletons, along with advances in robotics technology. Research on exoskeleton rehabilitation in the lower limbs has made continuous breakthroughs in various countries. Our country has gradually realized the importance of robots to the enhancement of scientific and technological strength, and has gradually begun to vigorously promote and develop the robot cause. In recent years, new robots have been emerging. Exoskeleton robots can help patients with leg rehabilitation training, which is more reproducible than traditional methods, and can greatly shorten the duration of rehabilitation training for patients. It also reduces the number of rehabilitative doctors accompanying them, greatly reduces the cost of human resources, plays a positive role in the rehabilitation of society and our country, and the development of robots, and provides effective assistance to the elderly and patients with lower extremity inconveniences in walking. This paper focuses on the rehabilitation exoskeleton humanoid gait because the gait is different from that of natural person. This gait is also called abnormal gait. First of all, the safety of gait is studied, the exoskeleton running state is judged by sensors, and the accidents occurring in walking are judged in real time. The exoskeleton robot used in this paper uses the default gait track. In order to improve the dynamic response of the drive system to the variable load condition, the joint load adaptive control is used in the drive joint. In the case of variable load, the servo drive system is able to follow the default gait in the control computer. This control method has better dynamic response characteristics than the traditional PID control mode. It can effectively improve the service life of servo system and mechanical structure. By analyzing and studying the gait characteristics of normal people, this paper combines the characteristics of exoskeleton four-legged walking gait, and improves the ability of servo system to follow the preset gait. Through a large number of experiments, the ideal gait trajectory can be obtained, and two paraplegic patients are allowed to walk autonomously. During the experiment, the author has carried out a large number of abnormal gait experiments. It was found that the way the wearer used crutches directly affected the stability of gait, especially at the beginning of learning quadruped gait. It is verified that the dynamic gait has higher stability than the traditional fixed gait and the center of gravity swing is smaller and more labor-saving under the condition of unreasonable support of crutches.
【学位授予单位】：中国科学院深圳先进技术研究院
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP242

【相似文献】