Fuzzy sliding mode control of a finger of a humanoid robot hand

Abstract: The motion control problem for the finger of a humanoid robot hand is investigated. First, the index finger of the human hand is dynamically modelled as a kinematic chain of cylindrical links. During construction of the model, special attention is given to determining bone dimensions and masses that are similar to the real human hand. After the kinematic and dynamic analysis of the model, in order to ensure that the finger model tracks its desired trajectory during a closing motion, a fuzzy sliding mode controller is applied to the finger model. In this controller, a fuzzy logic algorithm is used in order to tune the control gain of the sliding mode controller; thus, an adaptive controller is obtained. Finally, numerical results, which include a performance comparison of the proposed fuzzy sliding mode controller and a conventional sliding mode controller, are presented. The results demonstrate that the proposed control method can be used to perform the desired motion task for humanoid robot hands efficiently.     

下肢假肢斜坡路况运动控制策略分析

为使下肢假肢在斜坡路况下配合人体健肢侧交互运动,在对下肢假肢穿戴者斜坡运动模式分析的基础上,制定针对斜坡路况的下肢运动轨迹规划方法,得到假肢膝关节的模拟运动轨迹.应用紧格式线性化方法解决下肢关节高度非线性问题的思路,采取无模型动态矩阵控制方法实现对模拟运动轨迹的追踪,结合反馈校正和滚动优化实现斜坡路况下膝关节角度的预测输出,并对控制系统的稳定性进行理论分析.仿真和运动控制平台的实验结果表明,假肢膝关节的斜坡运动轨迹可以紧紧跟随参考轨迹变化,假肢膝关节可与健肢侧协调平稳运动.     

人手食指运动学建模

通过分析人手解剖学结构和观察人手运动特性,提出如下观点：（1）食指指掌关节具有两个轴线垂直相交的旋转自由度,两轴线相对于手掌不同的排列顺序对应于两种不同的运动学模型,指掌关节只能是其中的一种;（2）食指指掌关节的两个自由度在运动极限位置具有相关性．通过实验测绘出人手食指的工作空间形状,将之与两种可能的食指运动学模型对应的工作空间相比较,判别出指掌关节的真实运动学结构,并验证了指掌关节极限位置上两个自由度运动相关性．在上述工作基础上建立了符合人手实际结构的手指运动学模型．     

A finger motion model for reach and grasp

This study aimed to develop a model that describes human finger motion for simulation of reach and grasp for selected objects and tasks. Finger joint angles and timing of their changes were measured for six subjects as they reached 20-40 cm and grasped cylindrical handles (1.3-10.2 cm D) of varying orientation (vertical/axial). The empirical results from multiple regression analyses served as inputs to allow a fourth order polynomial to predict motion of each finger joint. The proposed model showed good fit with observations, with high coefficients of determination from 0.54 to 1 and reasonable errors from 0.04° to 5.44° for all conditions considered. The proposed finger motion model was implemented in an existing kinematic hand model to employ a contact algorithm for refined prediction of grip posture and to illustrate its predictive power by graphically displaying the opening and closing of the hand.

Relevance to industry

Finger joint motions during reach and grasp are needed for prediction of (1) tendon excursions for study of work-related musculoskeletal disorders, (2) required space for the hand, (3) finger locations on work objects, and (4) hand grip postures and strength.     

基于小波变换和样本熵的假手肌电模式识别方法

提出一种新的模式分类器,利用安置在拇长屈肌、指深屈肌和指伸肌上的3个电极所测得的肌电信号,实现了对3自由度假手手指运动的控制．该分类器采用小波变换和样本熵的方法构造特征矢量．经过由变学习速率算法和RP算法构建的集成3层前馈神经网络的分类,能够成功地分辨出拇指、食指和中指的弯曲与伸展运动,平均识别率可达96％以上．实验结果表明,该分类器为多自由度肌电假手的控制提供了一种有效的方法．     

仿生假手抓握力控制策略

为了使仿生假手完成各种精细作业,提出一种抓握力控制策略.在自由空间和约束空间中分别使用基于位置的阻抗控制和力跟踪阻抗控制.在过渡过程中使用模糊观测器切换控制模式.两种控制模式采用同一个基于位置的阻抗控制器,在约束空间向阻抗控制器中引入参考力,以满足约束空间的抓握力控制要求.这种方法可以使关节在自由空间和约束空间中分别实现良好的轨迹跟踪和力矩跟踪,在过渡过程中实现控制模式的可靠切换和系统的稳定过渡.提出一种自适应滑模摩擦力补偿方法,利用终端滑模思想设计了滑模函数,使得系统跟踪误差在有限时间内收敛,避免了传统线性滑模面状态跟踪误差无法在有限时间内收敛至0的问题.根据指数形式摩擦力的特点,利用终端滑模控制思想获得包含摩擦力参数估计的滑模控制律,并基于李亚普诺夫稳定性定理推导了估计参数的在线自适应律.对该抓握力控制策略在HIT假手上进行了抓取实验,实验结果证明了控制策略的有效性.     

新型多自由度假手的研制

介绍了一种新型多自由度假手的机械结构、运动方式及其静力学分析．假手的主要结构部件由光敏树 脂制成,使用小型直流伺服电机作为驱动器．手指采用欠驱动及腱驱动方式．该假手具有拟人的外观、较低的整体 重量和对被抓物体外形的适应能力．文章研究了假手的运动特点及接触力规律,通过仿真对假手的接触受力规律进 行了校验．最后通过实验,对假手的运动规律与抓握能力进行了验证．     

直线同步电动机磁悬浮系统的自适应模糊滑模控制

提出一种自适应模糊滑模控制方法用以提高可控励磁直线同步电动机(controllable excitation linear synchronous motor, CELSM)磁悬浮控制系统的性能.根据CELSM的特定结构和运行机理,建立CELSM磁悬浮系统的数学模型,包括励磁回路的电压方程、磁悬浮力方程和运动方程;设计积分滑模面和分段趋近律,系统状态轨迹可以根据距滑模面的距离自动切换趋近速度,以很小的斜率穿越滑模面,减小系统的抖振,推导出相应的滑模控制器;为了克服不确定性扰动对系统的影响,设计自适应律使自适应模糊系统对不确定性扰动进行实时估计,用该估计值进行前馈补偿控制,减小控制律中的切换增益和系统的抖振,进一步推导出自适应模糊滑模控制器;用Matlab对控制系统进行仿真,仿真结果表明,采用自适应模糊滑模控制的CELSM磁悬浮系统的性能得到改善.     

Fuzzy control design for the trajectory tracking in phase plane

Based on the concept of sliding mode control, the paper designs a fuzzy logic control for a linear system to achieve the trajectory tracking in the phase plane. That is, the system's state (e, e˙) is controlled to track the prespecified trajectory which is composed of several nonisoclinal segments. Each segment of the prespecified trajectory denotes the relation between the tracking error e and the error change e˙ in the corresponding region of the phase plane. In the paper, the prespecified trajectory is regarded as a sliding surface. A direct piecewise linear output formulation of the fuzzy logic controller is synthesized to achieve the hitting motion such that the trajectory tracking is completed region by region. We summarize the above analysis to be a fuzzy logic control design algorithm and give a practical example to illustrate its applicability     

基于Sigmoid函数的机器人鲁棒滑模跟踪控制系统设计

为了保证机器人能够在保持稳定的情况下,按照规划轨迹执行工作任务,从硬件和软件两个方面,设计了基于Sigmoid函数的机器人鲁棒滑模跟踪控制系统。装设机器人传感器与状态观测器,改装机器人鲁棒滑模跟踪控制器,完成系统硬件设计;综合机器人结构、运动机理和动力机制3个方面,构建机器人数学模型;根据状态数据采集结果与规划轨迹之间的偏差,计算机器人跟踪控制量;依据滑模运动与切换方程,利用Sigmoid函数生成机器人鲁棒滑模控制律,将生成控制指令作用在机器人执行元件上,实现系统的鲁棒滑模跟踪控制功能;在系统测试与分析中,所设计控制系统的平均位置跟踪控制误差为0.93 mm,与设定轨迹目标基本重合,机器人姿态角跟踪控制误差为0.06 mm,具有较好的鲁棒滑模跟踪控制效果,能够有效提高机器人鲁棒滑模跟踪控制精度。     

Rolling Contact Motion Generation and Control of Robotic Fingers

Dexterity in human hand is connected with the fingertip rolling ability. In this work we consider rolling motion of spherical robotic fingertips as one of the control objectives together with the set point position control and force trajectory tracking. The generation of a rolling motion trajectory is proposed and a control solution is designed which achieves prescribed transient and steady state tracking behavior. The proposed control law is structurally and computationally simple and does not utilize the dynamics of the robot model or its approximation. A simulation of a five degrees of freedom robot show excellent contact rolling performance even at cases of adverse friction conditions while alternative controllers lead to contact sliding. Experiments with a KUKA LWR4 + are performed to validate the proposed method.     

机械手的模糊逆模型鲁棒控制

提出一种基于模糊聚类和滑动模控制的模糊逆模型控制方法,并将其应用于动力学 方程未知的机械手轨迹控制.首先,采用C均值聚类算法构造两关节机械手的高木-关野 (T-S)模糊模型,并由此构造模糊系统的逆模型.然后,在提出的模糊逆模型控制结构中, 离散时间滑动模控制和时延控制(TDC)用于补偿模糊建模误差和外扰动,保证系统的全局 稳定性并改进其动态和稳态性能.系统的稳定性和轨迹误差的收敛性可以通过稳定性定理来 证明.最后,以两关节机械手的轨迹跟随控制为例,揭示了该设计方法的控制性能.     

基于位置和运动轨迹的老年人异常行为检测

针对老年人异常行为检测,分析老年人日常行为特点及规律,对其日常运动轨迹进行采样和量化,并使用基于改进的HMM模型进行轨迹行为习惯建模,结合模糊C均值聚类算法提取轨迹标志点确定HMM初值,并通过改进的重估公式和反馈滑动窗口进行训练与检测.实验结果表明,运动轨迹可以作为老年人行为习惯的一个重要描述形式,利用位置和轨迹对家庭环境下的异常行为检测是有效的,且能保证较高的准确率和较低的漏检率.     

基于数据手套与虚拟模型的手臂动作识别方法

为了配合残疾人使用助残假手完成日常活动,提出了一种基于数据手套与虚拟模型的手臂动作识别方法.通过数据手套上的MEMS传感器测量手臂运动数据,结合虚拟交互环境中的手臂模型,完成模型对真实手臂的动作复现,并获取手臂模型的位置数据.而后将传感器加速度、角速度等数据的时域特征与手臂模型位置数据的动作轨迹特征结合起来,使用支持向量机算法对预定义的5种手臂动作进行分类识别,识别率达到99.33％,验证了提出的手臂动作识别方法的可行性.     

具有感觉反馈的仿生假手人机交互研究

仿生假手是一个典型的生机电一体化系统.针对现有假手缺乏感觉反馈的问题,研究了具有双向信息传输能力的生机电一体化假手人机交互系统.采用压力传感器阵列检测肌肉压力分布信号,通过压力分布图的构建,基于主成分分析的特征提取和基于KNN分类器的特征识别方法实现了人手多运动模式的解码.基于假手力传感器,采用电刺激方式实现了假肢接触力向人体的感觉反馈.实验表明,基于肌肉压力分布信号的手部多运动模式识别方法具有较高的正确率,采用具有感觉反馈的交互控制系统可提高假手的抓取能力.     

Adaptive B-spline-based fuzzy sliding-mode control for an auto-warehousing crane system

In this paper, an adaptive B-splined-based fuzzy sliding mode control (ABFSM) is presented to the application of auto-warehousing crane motion control. The ABFSM comprises an adaptive fuzzy identification controller (AFIC) and a B-spline-based compensation controller (BCC). The AFIC is designed to approximate the ideal controller of a crane system. To alleviate the load of fuzzy rule base construction, only the information from the sliding surface is used as the input of AFIC such that the conciseness and translucency of the control system can be upgraded. On the other hand, the BCC aims to compensate the approximation error of the AFIC. With the introduction of the B-spline function, the boundary of the approximation error can be represented by means of polynomial mapping. Thus, the design of the compensation controller can be achieved based on the characteristics of the B-spline function. In this paper, the objective of the ABFSM is to track the distance-speed reference trajectory of the crane control system. With the tuning law of the AFIC and the BCC, the stability can also be guaranteed by means of Lyapunov function. To validate the performance of the proposed approach, the ABFSM is applied to auto-warehousing crane motion control under various conditions for x, y, and z directions, respectively. From the simulation the advantages of the proposed ABFSM are demonstrated, where the capability to handle the uncertainty with efficiency is verified.     

Fuzzy control design for the trajectory tracking on uncertainnonlinear systems

Based on the concept of sliding-mode control (SMC), the paper designs fuzzy logic controls to achieve the prespecified trajectory tracking for an uncertain nonlinear system. The prespecified trajectory is composed of several nonisoclinal segments in the phase plane and is regarded as the piecewise sliding surface. First, let the uncertain system be approximated by a linguistic fuzzy rule base, then two fuzzy logic controllers are designed to achieve the hitting motion and preserve the system's state traveling on the prespecified trajectory. The main advantage of this control design is that the trial and error of the conventional fuzzy control design disappears. A practical example is given to illustrate the applicability of the algorithm     

永磁球形电机的少保守性滑模控制

永磁球形电机轨迹跟踪控制方法常常利用高增益的控制输出来保证系统的鲁棒性及跟踪控制的快速性.但这种保守控制会带来较大的控制作用,甚至导致执行器饱和.为了减少控制的保守性,本文设计了一种带有非线性干扰观测器的模糊滑模控制器来解决球形电机的轨迹跟踪问题.利用干扰观测器对不确定性、摩擦、外界干扰、负载扰动等进行估计,并在控制输入端进行补偿实现对干扰的抑制.并利用滑模控制器抵消干扰观测器的干扰观测误差及不可观测部分的干扰,为了减少滑模的抖振,本文利用模糊逻辑对该部分进行逼近,并利用模糊的输出增益代替滑模的切换增益.此外通过Lyapunov方程证明了本文控制器的稳定性.仿真结果表明在存在模型不确定性及各种干扰的情况下,本文的轨迹跟踪控制具有良好的动静态性能和少保守性.     

Minimum jerk motion planning for a prosthetic finger

In this paper we propose a method, based on both physiologic and engineering considerations, for the motion planning of a prosthetic finger. In particular, we exploit a minimum jerk approach to define the trajectory in the Cartesian space. Then, cubic splines are adopted in the joint space. The redundancy problem arising from the presence of three links is solved by assuming that there is a constant ratio between the second and the third joint motion. The value of the proportional constant is determined by minimizing the maximum jerk in the joint space. It is found that this constant value can be suboptimally but effectively set to one for all the movements. This approach guarantees a natural movement of the finger as well as reduced vibrations in the mechanical structure and increased control performances. © 2004 Wiley Periodicals, Inc.     

Adaptive sliding mode control with moving sliding surface

A continuous sliding mode control with moving sliding surface for nonlinear systems of arbitrary order is presented in this paper. The sliding surface is moved repetitively toward the target sliding surface in order to ensure that the system trajectory is close to the actual surface during the whole control process. The parameters of sliding mode control are tuned by a fuzzy logic. The proposed procedure reduces the time when the system operates in the approaching phase during which the control performance is deteriorated since the system is more susceptible to external disturbances and model uncertainties. The effectiveness of the presented approach is demonstrated on a control of a flexible robot manipulator arm.