【作者】 李文凯；

【导师】 辛桂阳；

【作者基本信息】 大连理工大学 ， 光电信息工程， 2025， 硕士

【摘要】 随着社会医疗水平的提升,机器人技术在医疗康复领域的应用日益广泛,成为提升康复人群幸福度的重要手段。其中,康复机械臂因其较高的灵活性,在辅助上肢患者完成日常康复活动方面展现出显著优势。我国康复人群基数大,对智能化、个性化的康复辅助设备需求迫切。然而,现有上肢康复机械臂在运动控制方面仍存在适应性不足、依赖人工经验调整、轨迹规划效率低等问题,难以满足不同康复阶段患者的个性化需求。同时,康复机械臂的运动性能评价体系尚不完善,缺乏客观、量化的标准,影响了康复效果的评估。此外,构建兼顾康复指标和康复患者主观感受的综合评价体系,也是当前康复评价体系的重要挑战。本文以四自由度的康复机械臂为研究目标,为了促进上肢康复的整体过程更加科学规范,评价指标方案更加全面、合理,主要进行了康复机械臂的轨迹规划、轨迹评价方法及康复评价的研究。以下是本文的具体研究工作:基于D-H参数法对康复机械臂进行正向运动学理论模型推导,使用雅克比伪逆方法对康复机械臂进行逆运动学求解。制作了GUI显示界面,实现了随机点位的正逆运动学的求解,并利用蒙特卡洛数值计算法求得了整个系统的工作空间。针对上肢康复机械臂的轨迹规划问题,同时兼顾系统整体的运动随时间的变化符合康复者的需求,在笛卡尔空间中生成路径途经点,在关节空间中对轨迹进行设计规划,同时对末端构件的位置、速度和加速度情况进行显示。结果表明各项指标随时间变化图线平滑,保证上肢康复机械臂轨迹的稳定与可靠。围绕康复者在康复训练中的舒适感问题,对上肢康复机械臂的轨迹进行了优化。结合多项式对初始、终止和过程阶段分别进行了轨迹的设计优化,使得整个运动过程速度和加速度图线均具有匀速部分,并绘制了速度,加速度随时间变化的图像。引入了智能算法评价实际轨迹,从多种算法中挑选出具有代表性的10种方法,并进行了单峰测试和多峰测试,最终筛选出随机数评价方法并基于此方法对实际的轨迹进行评价。关于上肢康复过程评价问题,纳入10项指标进行多维度考虑(包括关节可动角度、肌肉紧张度、感觉系统功能、Brunnstrom评定、肌肉发力强度、自理能力、工具操作能力、移动辅助能力、社交互动运动能力以及文娱活动参与能力这十个评价指标)。通过R型聚类分析,将其分为两大类:硬性指标体系和软性指标体系,对硬性指标体系进行指标标准化,得到0-10标度的数值评价表。使用群体决策法构成判断矩阵,进行一致性检验合格后,将其带入到层次分析法中,并由此推出这五项指标在硬性指标体系中各自的权重。在软性指标体系中,由于熵权法确定权值含有对数运算,但原始数值标准化后可能包含零值或近似零值,为了保留原始数据中的信息,提出数值补偿法,并与去零法进行对比。将其应用于熵权法的计算中,得出各项指标的信息熵值和信息效用值,并计算出这五项指标在软性指标体系中的权重。更多还原

【Abstract】 With the improvement of social medical standards,the application of robot technology in the field of medical rehabilitation is becoming increasingly widespread,and it has become an important means to enhance the happiness of the rehabilitated population.Among them,the rehabilitation robotic arm has shown significant advantages in assisting upper limb pa-tients to complete daily rehabilitation activities due to its high flexibility.The base of the rehabilitation population in our country is large,and there is an urgent demand for intelligent and personalized rehabilitation assistive devices.However,the existing upper limb reha-bilitation robotic arms still have problems such as insufficient adaptability in motion control,reliance on manual experience for adjustment,and low efficiency in trajectory planning,mak-ing it difficult to meet the individualized needs of patients at different rehabilitation stages.Meanwhile,the evaluation system for the motion performance of rehabilitation robotic arms is still not perfect,lacking objective and quantitative standards,which affects the assessment of rehabilitation effects.Furthermore,constructing a comprehensive evaluation system that takes into account both rehabilitation indicators and the subjective feelings of rehabilitation patients is also an important challenge for the current rehabilitation evaluation system.This paper takes the four-degree-of-freedom rehabilitation robotic arm as the research goal.In order to promote the overall process of upper limb rehabilitation to be more scientific and standardized,and the evaluation index scheme to be more comprehensive and reasonable,the research mainly focuses on the trajectory planning,trajectory evaluation methods and re-habilitation evaluation of the rehabilitation robotic arm.The following is the specific research work of this article:The theoretical model of forward kinematics of the rehabilitation robotic arm is derived based on the D-H parameter method,and the inverse kinematics of the rehabilitation robotic arm is solved using the Jacobi pseudo-inverse method.A GUI display interface was created,the forward and inverse kinematics of random points were solved,and the workspace of the entire system was obtained by using the Monte Carlo numerical calculation method.Aiming at the trajectory planning problem of the upper limb rehabilitation robotic arm,while taking into account the changes of the overall movement of the system over time to meet the needs of the rehabilitated patients,the path passing points are generated in the Cartesian space,the trajectory is designed and planned in the joint space,and the position,velocity and acceleration of the end components are displayed at the same time.The results show that the graphs of various indicators changing over time are smooth,ensuring the stability and reliability of the trajectory of the upper limb rehabilitation robotic arm.Centering on the issue of comfort for rehabilitants during rehabilitation training,the tra-jectory of the upper limb rehabilitation robotic arm was optimized.Combining polynomials,the trajectories were designed and optimized respectively for the initial,final and process stages,so that both the velocity and acceleration graphs of the entire motion process had uniform parts,and the graphs of velocity and acceleration varying with time were drawn.Intelligent algorithms were introduced to evaluate the actual trajectories.Ten representative methods were selected from multiple algorithms,and unimodal and multimodal tests were conducted.Finally,the random number evaluation method was screened out and the actual trajectories were evaluated based on this method.Regarding the evaluation of the upper limb rehabilitation process,ten indicators were in-cluded for multi-dimensional consideration(including ten evaluation indicators such as joint range of motion,muscle tension,sensory system function,Brunnstrom assessment,mus-cle exertion intensity,self-care ability,tool operation ability,mobility assistance ability,so-cial interaction motor ability,and participation ability in cultural and recreational activities).Through R-type cluster analysis,it is divided into two major categories:the hard index system and the soft index system.The indicators of the hard index system are standardized to obtain the numerical evaluation table with a scale of 0-10.The group decision-making method is used to form a judgment matrix.After passing the consistency test,it is brought into the An-alytic Hierarchy Process,and from this,the respective weights of these five indicators in the hard indicator system are derived.In the soft index system,since the entropy weight method for determining weights involves logarithmic operations,but the original values may contain zero values or approximately zero values after standardization,in order to retain the informa-tion in the original data,a numerical compensation method is proposed and compared with the zeroing method.It is applied to the calculation of the entropy weight method to obtain the information entropy value and information utility value of each index,and calculate the weights of these five indicators in the soft index system.更多还原