【作者】 冯博；

【导师】 叶丹；

【作者基本信息】 东北大学 ， 控制理论与控制工程， 2020， 硕士

【摘要】 进入21世纪后,我国在各个领域都发生了巨大飞跃,人口老龄化趋势已经愈加明显,尤其近几年,已经逐步迈入老龄化社会。随着老年人人口的剧增,随之而来的将是各种各样的老年人疾病,尤其是神经类疾病以及脑血管疾病,都将造成肢体运动功能障碍。针对于这些病症目前最主要的治疗方法是由康复医师进行专业的康复训练。但是康复医师的治疗比较单一并且有很大缺陷,因此下肢外骨骼康复机器人技术近几年迅速发展,该技术能够帮助患者进行康复训练,是机器人技术与传统康复医学相结合的产物。对于偏瘫患者而言,病情分为多个不同的阶段,在不同的阶段需要采用不同的康复训练方式训练,因此康复训练也分为两种方式以对应于不同的阶段,分别为被动康复训练和主动康复训练。康复前期,患者基本不能自主活动,这时机器人带动患者进行训练。在康复中后期,患者肢体已具备活动能力,患者在进行康复训练时下肢外骨骼机器人起协助作用。权威研究表明,当患者主动进行康复训练时,受到损伤的神经系统能更快速的恢复。本文提出多种康复策略以应对患者的不同康复阶段。本文首先搭建了下肢外骨骼康复机器人系统,构建了以QNX系统为基础的实时系统框架,并分别搭建了地面实验平台以及跑步机实验平台。之后运用D-H参数法和拉格朗日法建立外骨骼运动学和动力学模型,之后对所建模型的正确性进行仿真验证,得到了机器人末端关节的姿态和各个关节角度的转换关系,为之后的控制打下了坚实的理论依基础。采用VICON光学运动捕捉系统采集正常人体步态,拟合并进行滤波处理,最终获得步态参考轨迹。获得步态轨迹后,针对于康复初期,采用PID步态轨迹跟踪被动控制进行被动训练。到了康复中后期,在传统导纳控制的基础上,本文提出了一种新的基于模糊自适应导纳控制的主动康复训练算法进行训练。当患者在进行康复训练时,根据人机交互力信息,控制器会在线调整导纳参数,以提高患者主动康复训练效果。最后,选用正常成年人进行了下肢外骨骼康复机器人实验,对于文中所提到的多种控制算法分别进行了验证,并进行了结果对比,验证了本文所使用的算法的可行性。又进行了外骨骼步态训练有效性实验,本文所搭建的外骨骼基本能够满足预期的效果。更多还原

【Abstract】 After entering the 21st century,China has made tremendous leaps in various fields,and the population aging trend has become more and more obvious,especially in recent years,it has gradually entered an aging society.With the rapid increase in the elderly population,various elderly diseases,especially neurological diseases and cerebrovascular diseases,will follow,which will cause limb motor dysfunction.At present,the main treatment method for these conditions is professional rehabilitation training by rehabilitation doctors.However,the rehabilitation doctor’s treatment is relatively simple and has great defects.Therefore,the lower extremity exoskeleton rehabilitation robot technology has developed rapidly in recent years.This technology can help patients to undergo rehabilitation training.It is a combination of robot technology and traditional rehabilitation medicine.For patients with hemiplegia,the condition is divided into multiple different stages,and different rehabilitation training methods are required at different stages.Therefore,rehabilitation training is also divided into two ways to correspond to different stages,which are passive rehabilitation training and Active rehabilitation training.In the early stage of rehabilitation,the patient was basically unable to move autonomously.At this time,the robot led the patient to train.In the middle and late stages of rehabilitation,the patient’s limbs are already capable of movement,and the patient’s lower extremity exoskeleton robot assists in the rehabilitation training.Authoritative research shows that when patients take the initiative to perform rehabilitation training,the injured nervous system can recover more quickly.This article proposes a variety of rehabilitation strategies to cope with the different rehabilitation stages of patients.This article first builds a lower limb exoskeleton rehabilitation robot system,builds a real-time system framework based on the QNX system,and sets up a ground experiment platform and a treadmill experiment platform,respectively.Later,DH parameter method and Lagrangian method were used to establish exoskeleton kinematics and dynamics model,and then the correctness of the model was verified by simulation.The transformation relationship between the posture of the robot end joints and the angles of each joint was obtained.Laid a solid theoretical foundation.VICON optical motion capture system was used to collect normal human gait,fit and filter it,and finally obtain the gait reference trajectory.After obtaining the gait trajectory,for the early stage of rehabilitation,passive gait trajectory tracking passive control was used for passive training.In the middle and late stages of rehabilitation,based on traditional admittance control,a new active rehabilitation training algorithm based on fuzzy adaptive admittance control is proposed for training.When the patient is undergoing rehabilitation training,the controller will adjust the admittance parameters online according to the human-computer interaction force information to improve the patient’s active rehabilitation training effect.Finally,experiments were performed on lower extremity exoskeleton rehabilitation robots using normal adults.The various control algorithms mentioned in the thesis were verified separately,and the results were compared to verify the feasibility of the algorithm used in this thesis.An experiment on the effectiveness of exoskeleton gait training was also carried out.The exoskeleton constructed in this thesis can basically meet the expected results.更多还原