基于神经网络的电机摩擦力混合模型研究

在电机伺服系统优化建模的研究中,要求高精度伺服系统。由于系统摩擦力具有强非线性和非光滑特性,传统的神经网络无法进行有效辨识。将非线性摩擦特性理解成为由稳态部分和突变部分串联构成,以电机伺服系统为对象,引入柔性sigmoid函数描述非线性摩擦特性中的突变部分,并与传统的RBF神经网络串联,构造出描述非线性摩擦特性的神经网络混合模型。仿真结果表明,与传统的RBF神经网络辨识方法相比,模型在输入变化响应下均具有较高的模型精度,从而验证了建模方法的有效性。     

动态迟滞Hammerstein系统的改进卡尔曼状态估计

迟滞特性具有非光滑、多值映射等复杂特性.而在实际的工程中,当输入电压变化频率超过一定的范围时,迟滞的特性是随着输入频率的改变发生变化,使得整个系统的状态估计工作更复杂.本文首先提出一种新的描述动态迟滞的方法,进而描述了动态迟滞Hammerstein系统的状态空间方程,根据此系统在传统卡尔曼滤波器的基础上进行改进得到一种新的非光滑卡尔曼滤波器.最后通过仿真和实验,比较了在输入信号变化频率比较大时,用动态迟滞Hammerstein系统来描述压电陶瓷和采用静态迟滞Hammerstein系统来描述压电陶瓷的特性,非光滑卡尔曼滤波器对这两种含有噪声的模型进行滤波,结果表明由于静态迟滞Hammerstein系统的建模不能很好的描述压电陶瓷的特性,模型存在着误差,因此对系统状态估计的结果也没有用动态Hammerstein系统的误差小,从而说明当输入电压频率变化比较大时研究动态的迟滞Hammerstein模型是很有意义的.     

基于构建类迟滞的高频响音圈电机建模与验证

在高频响应下,音圈电机存在明显的迟滞现象,并且具有一种特殊的复杂迟滞特性,即具有动态、非光滑和非单调的复杂迟滞特性.为了有效地描述复杂迟滞特性,提出了基于构建类迟滞的高频响音圈电机复杂迟滞混合神经网络建模方法,可实现复杂迟滞模型参数的自适应调整.实验结果表明,迟滞混合建模方法能有效地描述高频响音圈电机复杂迟滞特性以及对应的复杂迟滞的逆,并且模型具有较高的精度.     

压电陶瓷执行器的动态迟滞非线性特性建模

针对精密定位系统中压电陶瓷执行器的迟滞非线性特性建模问题,提出了一种基于Hammerstein迟滞模型的建模方法。通过引入一个Backlash类的算子来描述迟滞非线性的轮廓。在利用"扩展输入空间法"将迟滞特性的多值映射转换为一一映射的基础上,采用引力搜索算法优化的支持向量回归机建立静态迟滞模型。为体现迟滞的动态特性,用ARX模型表征迟滞环的率相关性,从而建立了Hammerstein级联模型。并从精密定位系统中采集了实测数据,通过电容传感器获取压电陶瓷执行器给定电压下的位移值,对所提出的模型进行了实验。实验表明:该模型具有较好的性能,辨识过程简便且易于工程实现。     

AFM压电陶瓷驱动器类Hammerstein建模与参数辨识

AFM( Atomic Force Microscope,原子力显微镜)中的压电陶瓷驱动器具有率相关迟滞非线性特性,这会影响AFM的扫描和定位精度。针对传统静态迟滞模型不能反映系统率相关动态迟滞特性的缺陷,提出Hammerstein模型以描述压电陶瓷驱动器的静态和动态迟滞特性。利用最小二乘支持向量机结合奇异值分解法对模型中的参数进行辨识。实验结果表明,模型能体现压电陶瓷驱动器的率相关迟滞特性,精度高于传统静态迟滞模型,建模方法对此类系统具有较好的应用价值。     

基于Preisach模型的迟滞系统建模与控制

针对一种复杂的非线性系统一迟滞系统，研究了基于KP算子Preisach模型对迟滞系统进行建模的方法。利用Preisach模型与其边界线之间的映射关系，建立了容易在线更新的迟滞模型。基于Preisach模型进行迟滞非线性系统的控制，采用PID方法来控制一类带有未知非线性特性迟滞的单输入单输出非线性系统。对迟滞非线性系统的建模与控制进行的数值仿真研究结果表明，该迟滞非线性系统的建模和控制方法具有理论意义和应用价值。     

基于多项式拟合的压电陶瓷迟滞神经网络建模

在压电陶瓷致动器优化设计的研究中,针对压电陶瓷的迟滞非线性特性,提出了一种基于多项式拟合算法的神经网络建模方法.由于压电陶瓷驱动器的迟滞现象是一种多对多的映射关系,而传统的建模方法只能对一对一映射进行建模.为解决上述问题,在对压电陶瓷迟滞现象的形成原因和特点进行深入分析的基础上,采用多项式拟合和神经网络相结合的方法对压电陶瓷驱动器的迟滞现象进行建模.仿真结果表明,采用多项式拟合算法的神经网络建模克服了传统建模方法只能对迟滞曲线进行分段建模的局限性,且拟合精度比较高,神经网络正模型的拟合误差为1.45％,神经网络逆模型的拟合误差为1.16％.表明上述神经网络模型精确地反映了压电陶瓷的迟滞特性.     

基于GMS摩擦模型的机器人低速运动研究

为了降低摩擦对机器人低速运动性能的干扰,引入了能够全面描述低速阶段摩擦力特性的GMS摩擦模型,并且利用遗传算法进行参数辨识,采用低通滤波器滤除高频噪声。为了验证该模型的有效性,设计了以机械手关节作为实验平台的轨迹跟踪实验,将模型作为前馈补偿引入系统。通过与常用的Stribeck模型补偿的对比,证明了GMS模型能更完善地描述摩擦力的特性,能进一步改善机械手的轨迹跟踪性能。     

基于径向基神经网络的压电作动器建模与控制

针对压电作动器(piezoelectric actuator,PEA)的率相关迟滞非线性特性,构建了Hammerstein模型对压电作动器建模.采用径向基(radial basis function,RBF)神经网络模型表征迟滞非线性,利用自回归历遍模型(auto-regressive exogenous,ARX)表征频率的影响,并对模型参数进行了辨识.此模型可以在信号频率在1～300 Hz范围内时,较好地描述压电作动器的迟滞特性,建模相对误差为1.99%～4.08%.采用RBF神经网络前馈逆补偿控制,结合PI反馈的复合控制策略实现跟踪控制,控制误差小于2.98%,证明了控制策略的有效性.     

基于神经网络的精密运动系统运动平台辨识

针对精密运动定位系统中宏动平台具有死区与迟滞的复合特性,提出了一种特殊的神经网络结构,将通常用于逼近光滑系统的神经网络模型改进为可以描述非光滑非线性特性的模型,在模型结构中引入一种非光滑激励函数,并引入广义梯度改进麦夸特算法,以用其对精密运动系统的含有非光滑非线性的运动特性进行建模.在所设计的神经网络中,同时也采用了扩展辩识空间方法,首先将迟滞特性的多值映射变为一一映射,而且还证明了采用完备化的算子基对辩识逼近的必要性及其扩展辩识空间的途径.实际辨识结果表明,所提出的建模方法取得了令人满意的结果.     

含有摩擦补偿的全方位移动机器人自抗扰控制

本文针对全方位移动机器人轨迹追踪中的摩擦补偿问题,提出了一种改进的非线性自抗扰控制器.首先建立了含有经典静态摩擦模型的全方位移动机器人动力学模型.其次,基于该模型设计非线性控制器和线性扩张状态观测器并给出了系统的稳定性分析.通过将模型已知项加入线性扩张状态观测器中得到摩擦力的估计值,并将估计值用于非线性控制器中摩擦补偿部分.为减小摩擦力对机器人低速运动轨迹追踪控制的影响,非线性控制器采用变增益控制器进行轨迹追踪控制.最后通过仿真结果验证本文提出控制器的有效性.     

An integrated friction model structure with improved preslidingbehavior for accurate friction compensation

Presents a dynamical friction model structure which allows accurate modeling both in the sliding and the presliding regimes. Transition between these two regimes is accomplished without a switching function. The model incorporates a hysteresis function with nonlocal memory and arbitrary transition curves. These last aspects prove essential for modeling presliding friction that is encountered in real physical situations. The model as a whole can also handle the Stribeck effect and stick-slip behavior as has been demonstrated by validation on a KUKA IR 361 robot. In this sense, this model can be considered as more complete in comparison with others found in the literature. The general friction model allows modeling of individual friction systems through the identification of a set of parameters that determine the complete behavior of the system. In this way, the model structure has been used to identify the friction behavior of a linear slide as well as that of the above mentioned KUKA robot. The results of the latter identification have been consequently used for feedforward friction compensation to obtain the most accurate tracking     

An adaptive friction compensator for global tracking in robot manipulators

A novel adaptive friction compensator based on a dynamic model recently proposed in the literature is presented in this paper. The compensator ensures global position tracking when applied to an n degree of freedom robot manipulator perturbed by friction forces with only measurements of position and velocity, and all the system parameters (robot and friction model) unknown. Instrumental for the solution of the problem is the observation that friction compensation can be recasted as a disturbance rejection problem. The control signal is then designed in two steps, first a classical adaptive robot controller that (strictly) passifies the system, and then a relay-based outer-loop that rejects the disturbance.     

行波型超声电机基于神经网络的逆模型辨识

行波型超声电机的动态特性受定子压电陶瓷迟滞和接触层非线性摩擦力的影响,表现出复杂的多值映射特征.通过引入动态迟滞逆算子,将存在于超声波电机逆系统中的多值映射在新的扩张输入空间上,转换为一一映射;然后使用神经网络建立超声波电机的逆模型,对迟滞和非线性摩擦力的影响进行补偿.所建立的模型结构简单,可以在线调整适应电机参数的非线性变化.实验仿真结果验证了该方法的有效性.     

Kinematic calibration for collaborative robots on a mobile platform using motion capture system

For modern robotic applications that go beyond the typical industrial environment, absolute accuracy is one of the key properties that make this possible. There are several approaches in the literature to improve robot accuracy for a typical industrial robot mounted on a fixed frame. In contrast, there is no method to improve robot accuracy when the robot is mounted on a mobile base, which is typical for collaborative robots. Therefore, in this work, we proposed and analyzed two approaches to improve the absolute accuracy of the robot mounted on a mobile platform using an optical measurement system. The first approach is based on geometric operations used to calculate the rotation axes of each joint. This approach identifies all rotational axes, which allows the calculation of the Denavit–Hartenberg (DH) parameters and thus the complete kinematic model, including the position and orientation errors of the robot end-effector and the robot base. The second approach to parameter estimation is based on optimization using a set of joint positions and end-effector poses to find the optimal DH parameters. Since the robot is mounted on a mobile base that is not fixed, an optical measurement system was used to dynamically and simultaneously measure the position of the robot base and the end-effector. The performance of the two proposed methods was analyzed and validated on a 7-DoF Franka Emika Panda robot mounted on a mobile platform PAL Tiago-base. The results show a significant improvement in absolute accuracy for both proposed approaches. By using the proposed approach with the optical measurement system, we can easily automate the estimation of robot kinematic parameters with the aim of improving absolute accuracy, especially in applications that require high positioning accuracy.     

An adaptive friction compensator for global tracking in robot manipulators

A novel adaptive friction compensator based on a dynamic model recently proposed in the literature is presented in this paper. The compensator ensures global position tracking when applied to an n degree of freedom robot manipulator perturbed by friction forces with only measurements of position and velocity, and all the system parameters (robot and friction model) unknown. Instrumental for the solution of the problem is the observation that friction compensation can be recasted as a disturbance rejection problem. The control signal is then designed in two steps, first a classical adaptive robot controller that (strictly) passifies the system, and then a relay-based outer-loop that rejects the disturbance.     

Maneuverability modeling and trajectory tracking for fish robot

This study develops a 6-DOF mathematical model for a robotic fish that considers surge, sway, heave, roll, pitch, and yaw. The model considers the conditions of a fish swimming in ocean current perturbations similar to the ocean current perturbations of the slender-body autonomous underwater vehicles. For swimming and turning behaviors, a nonlinear, dynamic, carangiform locomotion model is derived by using a planar four-link model. A 2-DOF barycenter mechanism is proposed to provide body stabilization and to serve as an actuating device for active control design. A barycenter control scheme is developed to change the center of gravity of the robot fish body by moving balancing masses along two axes. The projected torque on x and y axes propel pitch and roll angles to the desired settings. A Stabilizing controller, fish-tail mechanism, rigid body dynamics, and kinematics are incorporated to enable the fish robot to move in three dimensional space. Simulation results have demonstrated maneuverability and control system performance of the developed controller which is proposed to conduct path tracking of the robot fish as it swims under current perturbations.     

Posture Maintenance Control of 2-Link Object By Nonprehensile Two-Cooperative-Arm Robot Without Compensating Friction

In this paper, a method to posture maintenance control of 2-link object by nonprehensile two-cooperative-arm robot without compensating friction is proposed. In details, a mathematical model of the 2-link object is firstly built. Based on the model, stable regions for holding motion of nonprehensile two-cooperative-arm robot are obtained while the 2-link object is kept stable on the robot arms with static friction. Among the obtained stable regions, the robust pairs of orientation angles of the 2-link object are found. Under the robust orientation angles, a feedback control system is designed to control the arms to maintain the 2-link object’s posture while it is being held or lifted up. Finally, experimental results are shown to verify the effectiveness of the proposed method.      

Fuzzy-enhanced Adaptive Control for Flexible Drive System with Friction Using Genetic Algorithms

When a mechatronic system is in slow speed motion, serious effect of nonlinear friction plays a key role in its control design. In this paper, a stable adaptive control for drive systems including transmission flexibility and friction, based on the Lyapunov stability theory, is first proposed. For ease of design, the friction is fictitiously assumed as an unknown disturbance in the derivation of the adaptive control law. Genetic algorithms are then suggested for learning the structure and parameters of the fuzzy-enhancing strategy for the adaptive control to improve system's transient performance and robustness with respect to uncertainty. The integrated fuzzy-enhanced adaptive control is well tested via computer simulations using the new complete dynamic friction model recently suggested by Canudas de Wit et al. for modeling the real friction phenomena. Much lower critical velocity of a flexible drive system that determines system's low-speed performance bound can be obtained using the proposed hybrid control strategy.