面向直流电机模型时变参数的在线辨识方法

在直流电机建模优化问题的研究中,由于直流电机的时变参数随现场工况的变化而变化,导致采用离线辨识方法无法估计直流电机的系统模型,针对直流电机系统参数的时变性,提出一种直流电机时变参数的在线辨识策略.依据直流电机的电势平衡方程和转矩平衡方程,得到直流电机的传递函数,转化为电动机角速度和电压之间的差分方程,然后,采用递推最小二乘算法进行在线辨识,从而实现直流电机系统模型在线辨识的目的.搭建直流电机在线辨识系统实验平台,采用改变直流电机载荷的方式,进行了在线实时辨识的实验,实验结果表明:改进方法辨识效果良好,较好地实现了直流电机在参数时变的情况下,辨识效果良好,辨识模型较好地跟踪系统参数的变化.     

上肢康复机器人实时安全控制

针对上肢辅助康复机器人临床使用中的安全性和平稳性问题,提出基于模糊逻辑的实时在线安全监测控制方法.机器人对患肢进行康复训练时,患肢状态对控制效果会产生影响;通过设计智能安全监控模糊控制器(SSFC)改善系统运动平稳性以及突发情况下的安全性.首先提取相关运动特征评估受训患肢状态稳定情况,安全监控模糊控制器智能实现正常扰动情况下的控制期望力调节以及突发情况下的紧急响应.其次通过基于位置的阻抗控制策略实现患肢与机器人末端的柔顺性.实验结果验证了该控制方法能够有效地实现康复机器人的安全性和平稳性.     

用模糊模型在线辨识非线性系统

讨论用模糊方法实现非线性系统在线辨识问题.首先给出了简化的模糊规则表达 方法及其相应的自适应模糊推理,在此基础上给出了模糊模型参数在线辨识算法.最后对 非线性模型进行在线辨识,验证了本文提出的模糊模型及其在线辨识算法.     

基于在线辨识的特征模型预测函数控制研究

工业生产过程中运用模型控制问题,对于传统的预测函数控制,一旦预测模型与实际生产过程失配就会造成控制系统的控制性能下降.由于传统的预测函数控制算法中的预测模型通常是离线获得的,因此模型易出现失配.即使采用在线辨识方法来获取预测模型,由于在线辨识需花费时间,影响系统的实时性.通过构建实时数据库的监控层,实时采集生产过程控制参数,在监控层利用在线辨识方法辨识出被控过程的特征模型,然后将辨识出的特征模型作为预测函数控制算法的预测模型传至过程控制回路.采用获得在线辨识预测模型,不会影响系统的实时性.实验结果表明,根据在线辨识的特征模型与实际被控过程的失配很小,并提高了系统的控制性能.     

包含执行器动力学的子空间预测动态控制分配方法

针对具有冗余执行机构的过驱动系统的最优控制分配问题, 基于数据驱动的子空间辨识方法和预测控制理论, 提出了一种考虑执行器动力学特性的动态控制分配新方法. 在考虑范数有界不确定性的在线子空间辨识的基础上, 对执行器动力学特性进行不确定性建模, 再结合预测控制理论进行动态控制分配. 从而将执行机构的动力学建模、控制量最优分配和执行机构控制律的设计包含在一个子系统框架内, 对执行机构的模型不确定性具有更好的鲁棒性. 最后给出仿真实例验证了算法的有效性.     

基于模糊逻辑的上肢康复机器人阻抗控制实验研究

针对机器人辅助患肢主动康复训练过程中辅助力／阻力不能随患肢病情实时调整的问题,提出了一种 新的模糊自适应阻抗力控制方法．该方法实时检测患肢与机器人之间的相互作用力,并进一步运用辨识算法实时估 计出患肢的病情状态;然后运用模糊阻抗控制器对两者之间的相互作用力进行实时调整,使得在患肢主动能力不 足时提供一定的辅助,而在其有能力完成动作时,实时调整阻力实现肌力训练．实验结果表明了该控制方法的有效 性．     

基于去噪优化的动力系统关键参数辨识算法

由于传统方法没有对模型噪声实现有效处理,导致辨识精度与辨识速度较低,为此提出基于调制函数法的动力学系统参数辨识算法.对模型噪声进行表示与替换,通过调制函数法构造辨识模型,通过递推算法对辨识模型进行处理以实现噪声的处理.为构造出与动力学系统模型相一致的参数辨识模型,以动力学系统模型为基础对参数辨识模型的神经网络拓扑结构进...     

基于减法聚类的多模型在线辨识算法

考虑到实际工业过程中复杂系统的工况变化往往具有不确定性的特点, 离线辨识的多模型系统难以自适应反映系统的非线性, 因此本文提出一种新的基于减法聚类的多模型在线辨识算法. 首先采用在线聚类算法辨识多模型系统中的局部模型个数与工况参数, 然后充分考虑聚类发生变化对局部模型参数辨识的影响, 给出相应的局部模型参数在线辨识算法. 最后以某电厂300MW锅炉--汽轮机的协调控制系统为对象, 采用上述辨识方法进行仿真研究, 结果验证了本文算法的有效性.     

空间机器人动力学参数在轨辨识方法

针对机器人动力学参数识别问题,提出了基于有限遥测信息的空间机器人动力学参数在轨辨识方法.首先深入分析了各动力学参数对系统角动量的影响程度,并进行了参数重组,然后依据空间机器人系统的角动量守恒原理,推导了动力学参数与系统角动量之间的映射关系,最后,利用Gauss-Seidel迭代方法获得辨识结果.仿真表明,该方法能够快速地辨识出相关的动力学参数,方法利用遥测获得的飞行数据不需要额外的测量设备或者操作程序,为空间机器人动力学参数在轨实时辨识提供了理论方法.     

l1鲁棒辨识:一种递推插值方法

针对l1鲁棒辨识不能有效利用试验数据和进行在线辨识的问题,提出了一种在线递推插值辨识方法.用几何方法描述试验信息,利用系统可行集与新的试验信息所构成的半空间的包含关系判断数据信息,有效地利用了试验数据,提高了辨识精度.同时提出了一种新的计算辨识误差紧界的方法.仿真结果表明了算法的有效性和可行性.     

Control system design of a 3-DOF upper limbs rehabilitation robot

This paper presents the control system design of a rehabilitation and training robot for the upper limbs. Based on a hierarchical structure, this control system allows the execution of sequence of switching control laws (position, force, impedance and force/impedance) corresponding to the required training configuration. A model-based nonlinear controller is used to impose the desired environment to the patient's arm. The knowledge of robot kinematics and dynamics is thus necessary to ensure haptic transparency and patient safety. The identification process of robot dynamics is emphasised and experimental identification results are given for the designed robot. The paper also presents a particular rehabilitation mode named Active-Assisted. Simulation results of this rehabilitation mode illustrate the potentialities of the overall control scheme, which can also be applied to other rehabilitation robots.     

基于神经网络的机器人轨迹鲁棒跟踪控制

在神经网络辨识的基础上 ,提出一种新的鲁棒迭代学习控制方法。该方法利用神经网络对非线性系统进行在线辨识 ,产生迭代学习控制算法的前馈作用 ,并与实时反馈控制相结合 ,实现连续轨迹跟踪控制。仿真结果表明 ,该方法能克服机器人系统动力学模型的不确定性和外部干扰 ,且以极少的学习次数和网络训练次数达到满意的跟踪控制要求 ,具有良好的鲁棒性和控制性能     

On-line training of recurrent neural networks with continuoustopology adaptation

This paper presents an online procedure for training dynamic neural networks with input-output recurrences whose topology is continuously adjusted to the complexity of the target system dynamics. This is accomplished by changing the number of the elements of the network hidden layer whenever the existing topology cannot capture the dynamics presented by the new data. The training mechanism is based on the suitably altered extended Kalman filter (EKF) algorithm which is simultaneously used for the network parameter adjustment and for its state estimation. The network consists of a single hidden layer with Gaussian radial basis functions (GRBF), and a linear output layer. The choice of the GRBF is induced by the requirements of the online learning. The latter implies the network architecture which permits only local influence of the new data point in order not to forget the previously learned dynamics. The continuous topology adaptation is implemented in our algorithm to avoid memory and computational problems of using a regular grid of GRBF'S which covers the network input space. Furthermore, we show that the resulting parameter increase can be handled "smoothly" without interfering with the already acquired information. If the target system dynamics are changing over time, we show that a suitable forgetting factor can be used to "unlearn" the no longer-relevant dynamics. The quality of the recurrent network training algorithm is demonstrated on the identification of nonlinear dynamic systems.     

Identification and decentralized adaptive control using dynamicalneural networks with application to robotic manipulators

Efficient implementation of a neural network-based strategy for the online adaptive control of complex dynamical systems characterized by an interconnection of several subsystems (possibly nonlinear) centers on the rapidity of the convergence of the training scheme used for learning the system dynamics. For illustration, in order to achieve a satisfactory control of a multijointed robotic manipulator during the execution of high speed trajectory tracking tasks, the highly nonlinear and coupled dynamics together with the variations in the parameters necessitate a fast updating of the control actions. For facilitating this requirement, a multilayer neural network structure that includes dynamical nodes in the hidden layer is proposed, and a supervised learning scheme that employs a simple distributed updating rule is used for the online identification and decentralized adaptive control. Important characteristic features of the resulting control scheme are discussed and a quantitative evaluation of its performance in the above illustrative example is given.     

基于PSoC和DP-PSO-SVM的便携式假肢控制系统

针对三自由度假肢阈值控制方式存在不直观、灵活性差的缺陷，控制效果较好的模式识别控制器存在便携性差、实用性差等问题，提出一种基于PSoC在线模式识别的肌电假肢控制系统设计方案.采用低功耗芯片PSoC作为主控制器，设计了一套便携式四通道sEMG （表面肌电信号）采集系统，采用双群粒子群优化算法改进的支持向量机（DP-PSO-SVM）构建分类识别器，并通过假肢驱动器实现假肢在线模式控制.实验结果表明：采用DP-PSO-SVM算法比采用标准粒子群SVM （PSO-SVM）算法构建的分类器识别精度提高4%，达到96.7%；该控制器对6种动作的在线识别率达到96.3%，且符合实时性要求.     

Adaptive Controller for Single-Link Flexible Manipulators Based on Algebraic Identification and Generalized Proportional Integral Control

In this paper, we propose a fast online closed-loop identification method combined with an output-feedback controller of the generalized proportional integral (GPI) type for the control of an uncertain flexible robotic arm with unknown mass at the tip, including a Coulomb friction term in the motor dynamics. A fast nonasymptotic algebraic identification method developed in continuous time is used to identify the unknown system parameter and update the designed certainty equivalence GPI controller. In order to verify this method, several informative simulations and experiments are shown.     

基于操作者表现的机器人遥操作方法

针对遥操作机器人,提出了一种操作者表现（PoT）的在线识别方法以及基于PoT的遥操作移动机器人控制框架.首先,通过分析遥操作者的EEG（脑电信号）获得5个PoT指标,并使用BP（反向传播）神经网络对其进行建模,从而实现对遥操作者表现的在线识别.随后,设计了一种基于PoT动态调节遥操作共享控制系统中的控制权重的策略.选取3名不同遥操作者进行了在线PoT识别及基于PoT的共享控制遥操作实验,实验结果证明该方法能够有效地在线识别PoT,同时基于PoT的控制框架提升了遥操作的效率和安全性.     

Modeling and Verification of a Bimanual-Coordinated Training System

Recently, much attention has been paid to the development of robots that support bilateral arm training in various patterns. However, traditional bimanual rehabilitation robots usually realized different training modes with the robot providing a corresponding force for the impaired limb or else achieved an active-assisted mode with the healthy limb providing an assisted force for the impaired one. This paper proposes a robot to support bimanual-coordinated training. Different training modes are realized with one limb providing a corresponding force for the other limb. Two upper limbs accomplish symmetric movements in each training mode. Motion tracking training in active-resisted and active-assisted modes was performed on 11 healthy subjects. After bimanual-coordinated training, position tracking precision was significantly improved. The preliminarily results confirmed the feasibility of the system for supporting healthy subjects in performing bimanual-coordinated training tasks and demonstrated the effectiveness of the system in improving bimanual-coordinated performance of healthy subjects. Such a system could be potentially useful for patients who are in need of motor function rehabilitation after incidents such as stroke.     

一类离散时变系统的在线无限脉冲响应滤波逆控制

针对一类离散时变系统,提出了一种基于自适应惯性权重合作粒子群（AIW—CPSO）算法的在线尢限脉冲响应（IIR）滤波自适应系统辨识方法,实现零极点实时跟踪的全匹配控制．IIR滤波器可解决有限脉冲响应（FIR）滤波器在辨识时变系统时因其相关矩阵的特征值会无规律变大而被迫离线训练的问题．同时义降低了在线训练所需的权值向量长度,提升了优化与建模效率．本文设计的白适应惯性权重合作粒子群（AIW—CPSO）算法可在传统卡讧子群优化（PSO）算法的基础上更好地解决因选用IIR滤波器所带来的全局优化问题．通过仿真分析可以看出,对十此类离散时变系统,基于在线AIW—CPSO—IIR滤波器的自适应逆控制方法可以快速有效的实现未知对象的在线建模,同时实时跟踪时变系统的特征值变化．