Visual feedback control of a robot in an unknown environment (learning control using neural networks)

In this paper, a visual feedback control approach based on neural networks is presented for a robot with a camera installed on its end-effector to trace an object in an unknown environment. First, the one-to-one mapping relations between the image feature domain of the object to the joint angle domain of the robot are derived. Second, a method is proposed to generate a desired trajectory of the robot by measuring the image feature parameters of the object. Third, a multilayer neural network is used for off-line learning of the mapping relations so as to produce on-line the reference inputs for the robot. Fourth, a learning controller based on a multilayer neural network is designed for realizing the visual feedback control of the robot. Last, the effectiveness of the present approach is verified by tracing a curved line using a 6-degrees-of-freedom robot with a CCD camera installed on its end-effector. The present approach does not necessitate the tedious calibration of the CCD camera and the complicated coordinate transformations. This revised version was published online in October 2004 with a correction to the issue number.     

Vision-based stabilization of nonholonomic mobile robots by integrating sliding-mode control and adaptive approach

Vision-based pose stabilization of nonholonomic mobile robots has received extensive attention.At present,most of the solutions of the problem do not take the robot dynamics into account in the controller design,so that these controllers are difficult to realize satisfactory control in practical application.Besides,many of the approaches suffer from the initial speed and torque jump which are not practical in the real world.Considering the kinematics and dynamics,a two-stage visual controller for solving the stabilization problem of a mobile robot is presented,applying the integration of adaptive control,sliding-mode control,and neural dynamics.In the first stage,an adaptive kinematic stabilization controller utilized to generate the command of velocity is developed based on Lyapunov theory.In the second stage,adopting the sliding-mode control approach,a dynamic controller with a variable speed function used to reduce the chattering is designed,which is utilized to generate the command of torque to make the actual velocity of the mobile robot asymptotically reach the desired velocity.Furthermore,to handle the speed and torque jump problems,the neural dynamics model is integrated into the above mentioned controllers.The stability of the proposed control system is analyzed by using Lyapunov theory.Finally,the simulation of the control law is implemented in perturbed case,and the results show that the control scheme can solve the stabilization problem effectively.The proposed control law can solve the speed and torque jump problems,overcome external disturbances,and provide a new solution for the vision-based stabilization of the mobile robot.     

APPROACH TO IMPROVEMENT OF ROBOT TRAJECTORY ACCURACY BY DYNAMIC COMPENSATION

Some dynamic factors, such as inertial forces and friction, may affect the robot trajectory accuracy. But these effects are not taken into account in robot motion control schemes. Dynamic control methods, on the other hand, require the dynamic model of robot and the implementation of new type controller. A method to improve robot trajectory accuracy by dynamic compensation in robot motion control system is proposed. The dynamic compensation is applied as an additional velocity feedforward and a multilayer neural network is employed to realize the robot inverse dynamics. The complicated dynamic parameter identification problem becomes a learning process of neural network connecting weights under supervision. The finite Fourier series is used to activate each actuator of robot joints for obtaining training samples. Robot control system, consisting of an industrial computer and a digital motion controller, is implemented. The system is of open architecture with velocity feedforward function. The proposed m     

A QP Artificial Neural Network inverse kinematic solution for accurate robot path control

In recent decades, Artificial Neural Networks (ANNs) have become the focus of considerable attention in many disciplines, including robot control, where they can be used to solve nonlinear control problems. One of these ANNs applications is that of the inverse kinematic problem, which is important in robot path planning. In this paper, a neural network is employed to analyse of inverse kinematics of PUMA 560 type robot. The neural network is designed to find exact kinematics of the robot. The neural network is a feedforward neural network (FNN). The FNN is trained with different types of learning algorithm for designing exact inverse model of the robot. The Unimation PUMA 560 is a robot with six degrees of freedom and rotational joints. Inverse neural network model of the robot is trained with different learning algorithms for finding exact model of the robot. From the simulation results, the proposed neural network has superior performance for modelling complex robot’s kinematics.     

基于递归小脑神经网络模型控制的空间机器人关节抗死区及摩擦控制

探讨了存在关节力矩输出死区、摩擦与外部干扰的载体位姿均不受控的漂浮基空间机器人系统的动力学控制问题。设计了一种递归小脑模型关节控制器（CMAC）神经网络与死区估计补偿器,使两关节铰能够跟踪期望运动轨迹。该控制器利用摩擦双观测器估计不可测的内部摩擦状态,利用死区预补偿器消除关节力矩输出死区的影响;应用递归小脑神经网络模型逼近了包含摩擦误差及外部干扰的动力学方程不确定项。仿真结果表明了该控制方法的有效性。     

Optimal trajectory planning with application to industrial robots

A procedure is presented for planning optimal trajectories for application to industrial robots. First, trajectories are optimised by considering the nominal dynamics of a robot with rigid links and joints and with constraints on joint torque and speed. The minimum-time optimisation criterion is complemented by a miminal dynamic energy criterion that leads to smoother actuator inputs that do not excite joint vibrations. Weighting factors for these cost functions are then determined by trial simulations. By these means the effect of controller characteristics and elasticity, friction and backlash in the joints may be taken into account. A minimum-time movement for the real-world robot is obtained which displays the dynamical behaviour predicted in the planning procedure. Results from measurements and simulations for a PUMA 562 robot illustrate the approach. Further improvements may be achieved by a custom controller with the feedforward torques as shown in a comparison of trajectories executed with a VAL2 controller and a custom controller.     

柔性和摩擦力不确定条件下RBF神经网络自适应轨迹跟踪方法

提出一种基于径向基神经网络(Radial basis function, RBF)的力/位置混合自适应控制方法并用于机器人轨迹跟踪控制,解决机器人柔性末端执行器轨迹跟踪过程中柔性和摩擦力模型难以精确描述的问题。RBF神经网络是一种高效的前馈式神经网络,具有其他前向网络所不具有的非线性逼近性能和全局最优特性,并且网络结构简单,训练速度快。设计一种基于RBF神经网络非线性逼近能力来估计模型中的不确定参数的自适应控制器,给出控制器中神经网络权值更新规则,并证明所设计控制器输出力和位置误差的最终一致有界性。将该控制器应用于风管清扫机器人仿真试验,结果表明该自适应控制器能很好地用于柔性和摩擦力不确定条件下轨迹跟踪控制,与传统自适应控制方法相比具有更精确的跟踪特性和更强的鲁棒性。     

气动人工肌肉外骨骼机器人位置跟踪控制

针对以气动人工肌肉作为关节驱动器的外骨骼机器人关节位置跟踪控制问题进行了研究。首先,在动力学模型的基础上,设计了上层控制器,并结合自适应控制和滑模控制方法降低了动力学参数不准确和扰动项未知对外骨骼机器人的影响;其次,基于无模型方法设计了底层关节力矩控制器,调整外骨骼机器人的关节力矩;最后,针对上述控制方案设计仿真实验与外骨骼机器人的穿戴实验。结果表明,该控制方法对气动人工肌肉外骨骼机器人的关节位置跟踪控制是有效的。     

一种新型的模糊神经网络控制器

本文在分析传统模糊控制的基础上,提出了一种由模糊神经网络实现的自适应模糊控制器,详细分析了网络的结构形式、控制器的在线学习、离线学习方法。最后成功地将它应用于两关节液压机器人的实时控制之中。     

空间机器人捕获航天器操作的避撞柔顺无源神经网络H∞控制

研究了空间机器人在轨捕获非合作航天器过程避免关节受冲击破坏的避撞柔顺控制问题。为此在关节电机与机械臂之间配置了一种柔顺机构--旋转型串联弹性执行器(RSEA),可通过其内置弹簧的变形来吸收捕获过程目标航天器对空间机器人关节产生的冲击能量;结合所设计的开、关机控制策略可保证关节冲击力矩受限在安全范围内。首先利用拉格朗日方法及牛顿-欧拉法分别获得了捕获前空间机器人及目标航天器的分体系统动力学模型;之后,结合冲量定理、系统运动几何关系及力的传递规律,建立了捕获后两者形成混合体系统的动力学模型,并计算了碰撞过程的冲击力矩;最后,基于无源性理论提出了一种神经网络鲁棒H∞避撞柔顺控制策略以实现失稳混合体的镇定控制。数值仿真结果表明,配置柔顺空间机器人在捕获碰撞阶段最大可减小61.9%的关节冲击力矩,最小也可减小47.8%;而在镇定运动阶段,各关节冲击力矩均受限在安全范围内,实现了对关节有效地保护。     

一种足式移动机器人的混合视觉伺服控制方法

针对一种配有单目相机的足式移动机器人,提出一种基于图像单应性的混合视觉伺服方法,利用单应性矩阵中的元素构造状态变量估计估计机器人相对位姿,使机器人在缺乏深度信息的情况下可准确到达目标位姿。相较于轮式移动机器人,在足式机器人移动过程中,足式机构的间歇性运动直接影响机器人视觉反馈过程和伺服控制系统的准确性和稳定性。为解决这一问题,通过分析足式机构运动学并建立足式机器人移动速度与电机转速的映射关系,使控制器可以更准确地调整机器人运动速度。考虑到足式机构运动对视觉反馈环节的影响,提出一种改进型自适应中值滤波算法提高位姿估计精度。伺服环节设计了滑模控制器,并采用李雅普诺夫方法证明了控制系统的稳定性。最后,利用CoppeliaSim软件搭建足式移动机器人虚拟模型,通过仿真验证了所提出控制方案的可行性与有效性。     

自由漂浮空间机械臂基于神经网络的鲁棒自适应控制

考虑数学模型难以精确获得及带外部干扰情况下,针对自由漂浮空间机械臂的轨迹跟踪控制问题,提出一种基于神经网络的自适应鲁棒控制策略。基于Lyapunov稳定性理论设计理想控制器,进而推出系统的不确定模型。利用神经网络的学习能力逼近系统不确定模型,从而避免保守上界的估计。利用线性化技术并结合Lyapunov函数,设计包括权值及隐层参数在内的在线自适应学习律及鲁棒控制器,加快了误差收敛速度及控制精度,并消除了高阶逼近误差及扰动,保证了系统的一致最终有界,仿真比较表明了该控制策略的有效性。     

神经滑模控制在转台伺服系统应用

以直流电机为执行机构,分析了飞行仿真转台伺服系统的数学模型。滑模控制具有对系统扰动和参数摄动的自适应性,可实现伺服系统的快速响应,同时有效克服低速状态下摩擦力矩的影响。系统抖振问题是滑模控制的突出问题,利用径向基神经网络的非线性逼近能力,给出以切换函数为网络输入,以滑模控制器为网络输出,构建了神经滑模控制器,软件仿真结果表明所设计的滑模控制器能达到较好的控制品质,有效的克服系统抖振和外部扰动,实现系统低速摩擦补偿。     

电液负载模拟器复合控制方法研究

提出了一种新的电液负载模拟器复合控制方案,采用改进型多余力补偿方法和PID自适应控制器并行的方式来实现对指令控制力的精确跟踪。改进型的前馈补偿方法除了有效地消除了多余力,还提高了系统的动态性能。利用CMAC神经网络的非线性逼近原理设计的鲁棒PID自适应控制器,在一定程度上改善了传统PID控制在快速性和稳定性之间存在的矛盾,降低了系统的非线性和不确定性造成的影响。仿真和试验结果证明了该控制方案的有效性。     

基于径向基函数网络的MOTOMAN机械手运动学逆解

从集合和数学观点 ,把运动学正解和逆解问题作为机器人关节空间和工作空间之间的非线性映射关系 ,将运动学逆解过程转换为神经网络权值训练问题。基于具有局部逼近能力的特点 ,将正解结果作为训练样本 ,用 6输入、单输出的RBF网络 ,实现了MOTOMAN机械手运动学逆解计算 ,避免了传统方法的繁琐公式推导。算例表明 ,采用RBF网络解决逆解问题比BP网络的计算精度略有提高。此外 ,RBF网络有更快的收敛速度     

基于模糊神经网络的不确定机器人实时轨迹跟踪控制的研究

针对机器人建模的不精确性以及扰动的存在给机器人控制增加难度的问题,提出了一种基于模糊神经网络的不确定机器人实时轨迹跟踪控制方法。该控制方法的控制器由模糊神经网络(FNN)控制器和CMAC控制器组成,FNN控制器代替传统的计算力矩法,CMAC控制器在线补偿控制误差,有效补偿机器人存在的各种不确定性。对二自由度机器人的仿真结果表明了所提出的控制方法的可行性。     

Implementation of robust adaptive control for robotic manipulator using TMS320C30

A new adaptive digital control scheme for the robotic manipulator is proposed in this paper. Digital signal processors are used in implementing real time adaptive control algorithms to provide an enhanced motion for robotic manipulators. In the proposed scheme, adaptation laws are derived from the improved Lyapunov second stability analysis based on the adaptive model reference control theory. The adaptive controller consists of the adaptive feedforward and feedback controller and PI type time-varying control elements. The control scheme is simple in structure, fast in computation, and suitable for implementation of real-time control. Moreover, this scheme does not require an accurate dynamic modeling, nor values of manipulator parameters and payload. Performance of the adaptive controller is illustrated by simulation and experimental results for a SCARA robot.     

机器人焊缝跟踪神经网络控制的研究

研究神经网络技术在弧焊机器人焊缝跟踪过程中的应用,通过神经网络在笛卡尔空间轨迹的补偿作用,确定出基于笛卡尔空间参考轨迹控制的机器人焊缝跟踪神经网络控制器。与传统的关节计算力矩法相比,所设计的神经网络控制器具有良好的控制特性及较强的鲁棒性,焊缝跟踪精度得到了显著的提高。     

Simulation and stability analysis of neural network based control scheme for switched linear systems

This paper proposes a new adaptive neural network based control scheme for switched linear systems with parametric uncertainty and external disturbance. A key feature of this scheme is that the prior information of the possible upper bound of the uncertainty is not required. A feedforward neural network is employed to learn this upper bound. The adaptive learning algorithm is derived from Lyapunov stability analysis so that the system response under arbitrary switching laws is guaranteed uniformly ultimately bounded. A comparative simulation study with robust controller given in [Zhang L, Lu Y, Chen Y, Mastorakis NE. Robust uniformly ultimate boundedness control for uncertain switched linear systems. Computers and Mathematics with Applications 2008; 56: 1709-14] is presented.     

Tracking control of a three-wheeled omnidirectional mobile manipulator system with disturbance and friction

This paper proposes a tracking control method for a three-wheeled omnidirectional manipulator system (OMMS) with disturbance and friction. The OMMS is separated into two subsystems, a three-wheeled omnidirectional mobile platform (OMP) and a selective compliant articulated robot for assembly (SCARA) type of manipulator. Therefore, two controllers are designed to control the OMP and the manipulator system. Firstly, based on a kinematic modeling of the manipulator, a kinematic controller (KC), combined with an integral sliding mode controller (ISMC), is designed for the end-effector of the manipulator to track a desired trajectory with the desired angular velocity vector of links. Secondly, a differential sliding mode controller (DSMC) based on a dynamic modeling of the OMP with force external disturbances is proposed to obtain control inputs moving the OMP so that the manipulator tracks the desired posture without singularity. The system stability is proven using Lyapunov stability theory. The simulation and experimental results are presented to illustrate the effectiveness of the proposed controllers in the presence of disturbance and friction.