自适应环境的机器人视觉伺服控制方法

机器人和机器视觉的迅速发展,使得基于视觉的智能机器人得到更加广泛的应用。机器视觉提高了机器人控制系统对环境的适应程度,但其适应程度也受到周围环境对机器视觉算法的影响。针对这一情况,本文对机器人视觉伺服控制系统中的机器视觉算法进行了改进,提出了一种基于帧间差分法的自适应环境的机器人视觉伺服控制方法,从而提高了控制系统对环境的自适应程度,提高了对机器人控制的精确度,实验和理论分析证明,该方法具有较大的应用前景和实用价值。     

Direct method for updating flexible multibody systems applied to a milling robot

Industrial robots are currently used in light milling operations for their low cost and large workspace compared with CNC machine tools. However, milling robots are prone to vibration instabilities (chatter) and process deviations since they are significantly less stiff than machine tools. As a result, robot dynamic response depends on its posture which represents a major challenge. This paper presents a direct method to update any multibody model, enclosing flexible rotational/translational or virtual joints with minimal tuning. The novel method allows determining the elastic parameters of the model based on a curve fitting of the frequency response functions measured at the tool tip. Fitting is fast and efficient as it occurs in the frequency domain without the need to transform the measured data into the model parameter space. It relies on a genetic algorithm followed by a deterministic procedure to ensure a refined solution of the identified global minimum. The method is firstly demonstrated and validated on a simulated flexible manipulator with three rotational joints. Its multibody model is built using minimal coordinates with known elastic parameters that the method recovers accurately. The new fitting algorithm is eventually applied to an actual industrial robot (KUKA KR90 R3100 robotic arm) resulting in the proper fit of its critical resonances. Posture dependency can also be tackled by considering multiple measurements in different poses within the same fitting procedure. Updating procedure was programmed in Matlab and made public so that it can be easily adapted to identify elastic parameters of other flexible mechanical systems.     

On the position analysis of a new spherical parallel robot with orientation applications

In this paper, we propose a new spherical parallel robot for celestial orientation, and rehabilitation applications (TV satellite dish, tracking systems, solar panels, cameras, telescopes, table of the machine tools, ankle, shoulder, wrist and etc.). The proposed robot can completely rotate about an axis. After describing the robot and its inverse position analysis, using the genetic algorithm, the dimensional optimization to maximize the workspace of the robot is performed. The workspace analysis shows that the proposed robot has a relatively large workspace. Also, singularity analysis represents that the manipulator is a singularity-free workspace. It is a great advantage of the proposed robot. Next, an optimal approach is proposed for solving the direct position problem of the robot. According to the geometry of the robot, two coupled trigonometric equations are obtained through using a special form of Rodrigues' rotation formula. Next, the two coupled equations are transformed to a 8-degrees polynomial using the Sylvester's Dialytic elimination method. Finally, a numerical example for the robot with an asymmetric structure is given with eight real solutions. Therefore, the polynomial being minimal and the proposed approach is optimal. This greatly decreases computational time, which is necessary for dynamics, control and simulation.     

仿人机器人步态平衡泛化模型的建立与仿真

通过人体示教计算零力矩点(zero moment point, ZMP),并通过补偿关节角度对其矫正的方法可以解决机器人步行不稳定的问题,但仍存在算法复杂度过高等问题。本文提出一种人体示教与机器学习相结合的方法,基于支持向量回归算法建立机器人的步态平衡泛化模型,通过该模型可以实现对模型输入人体示教的关节角度和ZMP信息后直接得到经稳定性补偿的关节角度,并以此驱动机器人完成步行动作。引入鲸鱼优化算法(whale optimization algorithm, WOA)优化模型的参数以使模型得到最优的泛化效果,完善步态平衡模型的性能。WEBOTS仿真平台下,使用模型输出的补偿后的关节角度驱动NAO机器人,其动作自然、稳定且算法复杂度较低,验证了本文方法的可行性。     

基于PSO-ELM的机器人精度补偿方法研究

为了提高机器人的定位精度,对传统的基于神经网络的机器人精度补偿方法进行改进。采用两种基于粒子群优化的极限学习机（PSO-ELM）模型的精度补偿方法对机器人关节坐标及直角坐标进行补偿。分别对两种方法进行仿真实例分析,并与遗传算法优化的极限学习机（GA-ELM）模型进行对比。仿真结果表明,对直角坐标进行补偿的PSO-ELM机器人精度补偿法优于其他补偿方法,且具有较高的预测精度。     

Iterative learning control for integrated system of robot and machine tool

This study is concerned with the integrated system of a robot and a machine tool. The major task of robot is loading the workpiece to the machine tool for contour cutting. An iterative learning control (ILC) algorithm is proposed to improve the accuracy of the finished product. The proposed ILC is to modify the input command of the next machining cycle for both robot and machine tool to iteratively enhance the output accuracy of the robot and machine tool. The modified command is computed based on the current tracking/contour error. For the ILC of the robot, tracking error is considered as the control objective to reduce the tracking error of motion path, in particular, the error at the endpoint. Meanwhile, for the ILC of the machine tool, contour error is considered as the control objective to improve the contouring accuracy, which determines the quality of machining. In view of the complicated contour error model, the equivalent contour error instead of the actual contour error is taken as the control objective in this study. One challenge for the integrated system is that there exists an initial state error for the machine tool dynamics, violating the basic assumption of ILC. It will be shown in this study that the effects of initial state error can be significantly reduced by the ILC of the robot. The proposed ILC algorithm is verified experimentally on an integrated system of commercial robot and machine tool. The experimental results show that the proposed ILC can achieve more than 90% of reduction on both the RMS tracking error of the robot and the RMS contour error of the machine tool within six learning iterations. The results clearly validate the effectiveness of the proposed ILC for the integrated system.     

改进灰狼优化算法求解机器人作业车间调度问题

针对物料机器人指派和作业车间的联合调度问题,设计了一种改进灰狼优化算法进行求解。根据机器人作业车间调度和灰狼优化算法的各自特点,提出一种面向机器人转移工序的编码方式。解码时,考虑工件运输的前提是工件在当前机器的工序已加工,提出融合间隙解码方法的驱动解码方法。为避免算法陷入局部最优,在灰狼个体位置更新后加入个体变异方法。最后,通过与其他智能优化算法及同类算法进行比较,验证了所提灰狼优化算法的有效性。     

Development of CAM system based on industrial robotic servo controller without using robot language

This paper describes the development of a robotic CAM system for an articulated industrial robot RV1A from the view point of robotic servo controller. It is defined here that the CAM system includes an important function which allows an industrial robot to move along cutter location data (CL data) consisting of position and orientation components. In addition, the developed CAM system has a high applicability to other industrial robots whose servo systems are technically opened to end-users. The developed robotic CAM system works as a straightforward interface between a general CAD/CAM and an industrial robot. At the present stage, the relationship between CAD/CAM and industrial robots is not well established compared to NC machine tools that are widely spread in manufacturing industries. The CAM systems for NC machine tools are already established, however, the CAM system for industrial robots has not been sufficiently considered and developed yet. A teaching pendant is generally used to obtain position and orientation data of the arm tip before an industrial robot works. Here, in order to enhance the relationship between a conventional CAD/CAM system and an industrial robot, a simple and straightforward CAM system without using any robot language is developed and implemented. The basic design of the robotic CAM system and the experimental results are presented in this paper.     

Collision avoidance for Robot manipulators: Application to Catia/IBM 7565 interface

A collision avoidance algorithm has been developed to augment the capability of an automatic (off-line) robot path planning (programming) tool. The use of off-line programming tools for robot task programming is becoming increasingly important, but the advantages to be gained by off-line programming may be lost if collision-free path planning capabilities are not included. This article addressed the problem of collision-free path planning in the context of a gantry type robot. The collision avoidance algorithm described here uses the <heuristic approach> to collision-free path planning. The manipulator and obstacles are modeled as spheres to simplify tests for collision. An important feature of this algorithm is that it permits the manipulation of objects in the robot's environment. When compared against an algorithm from the literature, given a lightly cluttered environment modelled by spheres, the new algorithm finds a collision-free path much faster. This new algorithm has been implemented as part of the CATIA/IBM 7565 interface which forms an automatic off-line programming system for the IBM 7565 robot. It has also been implemented as a supervisory collision filter to allow collision-free control of the robot from the operator's console. In both cases the algorithm has been demonstrated to provide efficient and effective collision avoidance for the IBM 7565 robot.     

基于视觉导航的轮式移动机器人横向最优控制

近年来，基于机器视觉导航的轮式移动机器人在农业领域中被广泛研究．本文运用 最优控制算法对其进行横向控制，但是，农田非结构化自然环境的机器视觉识别耗时导致反 馈通道传输延迟明显，最优控制器的性能因之而下降．因此，本文又根据轮式移动机器人运 动学模型预测其行为，修正了滞后的反馈信息．仿真结果表明，该算法能实现轮式移动机器 人精确可靠的横向控制，具有较强的纵向速度自适应能力．     

A noninverting algorithm for path tracking of two cooperating robot arms and its parallel implementation

This paper presents an algorithm for path tracking of two robot arms with end-effectors gripping a common inertial load. The path is generated as a sequence of elementary motions. The most important feature of the present algorithm is that it avoids singularities, because there is no need of using the inverse kinematics. Direction and proximity criteria are introduced. Holonomic constraints are formulated for the position and orientation of the two end-effectors.The application of parallel processing methods to path tracking according to the previous algorithm is presented. The algorithm is implemented in the Alliant FX/80 parallel machine.     

基于结构约束的架空输电线路巡线机器人障碍识别

巡线机器人沿相线行走时必须探测识别各种障碍，并根据障碍类型规划越障行为．针对220 kV架空输电线路的结构特点，利用视觉传感器，设计了基于结构约束的障碍识别算法．算法利用图像的边缘信息，采用改进的基于存在概率图的圆/椭圆检测方法和分层决策机制，以减少自然环境中光线变化和机器人运动对识别质量的影响，满足了巡线机器人的实时越障要求．实验室模拟线路和实际线路实验结果表明，算法能可靠地识别出复杂背景中的防震锤、悬垂线夹和耐张线夹等障碍物．     

基于LSSVM的静态手势识别

支持向量机(Support Vector Machine，简称SVM)，是基于统计学习理论的一种新的模式识别方法，较好地解决了小样本学习问题。通过使非线性空间变换为线性空间，降低了算法的复杂性。LSSVM(Least Squares Support Vector Machine)由于使用线性等式代替了标准的SVM算法中的线性不等式，进一步降低了运算量。利用傅立叶描述子获取静态手势特征向量，通过LSSVM大尺度算法求解方程组来得到LSSVM分类器，进行静态手势识别，取得了较高的识别率。说明如何把静态手势识别结果应用到机器人远程控制中，提高人机交互的友好性。     

稳态调度下单臂组合设备时间延迟分析与优化

在半导体晶圆制造过程中,驻留时间延迟过长对晶圆质量具有消极影响.本文研究单臂组合设备稳态调度中如何合理地分配机械手等待时间,抵消驻留时间延迟的问题.首先,采用Petri网模型描述晶圆制造过程,分析了单臂组合设备稳态调度的时间特性,获得了稳态下工序驻留时间延迟计算表达式.其次,通过解构机械手等待时间对驻留时间延迟的影响机理,提出了一种机械手等待时间分配优先级规则.进一步,将虚拟瓶颈工序用于辅助分配机械手等待时间,结合优先级规则,提出了一种单臂组合设备稳态调度启发式算法.最后,通过例子验证了算法的可行性与有效性.与传统拉式策略和尽早加工策略对比,该算法能有效地减少单臂组合设备稳态调度下的驻留时间延迟并能满足晶圆制造的严格要求.     

Path planning approach in unknown environment

This paper presents a new algorithm of path planning for mobile robots, which utilises the characteristics of the obstacle border and fuzzy logical reasoning. The environment topology or working space is described by the time-variable grid method that can be further described by the moving obstacles and the variation of path safety. Based on the algorithm, a new path planning approach for mobile robots in an unknown environment has been developed. The path planning approach can let a mobile robot find a safe path from the current position to the goal based on a sensor system. The two types of machine learning: advancing learning and exploitation learning or trial learning are explored, and both are applied to the learning of mobile robot path planning algorithm. Comparison with A* path planning approach and various simulation results are given to demonstrate the efficiency of the algorithm. This path planning approach can also be applied to computer games.     

冷凝管水下视觉定位系统研究

根据机器人水下工作环境和工况要求,针对冷凝管清洗设备坐标输入法定位不准的问题引入了机器视觉技术,分析了系统的可行性设计,设计了一套基于DSP的嵌入式机器视觉系统,并详细叙述了照明方式、图像采集模块、图像处理模块的组成及工作原理,最后结合系统应用要求和算法的实时性,通过最大方差法及重心法完成了管口圆心的定位。该视觉系统取得良好的实验结果,能在一定程度上提高整个机器人系统的集成度、定位的快速性与精确度。     

Predictive pull-based control of an unmanned manufacturing cell, accounting for robot mobility

In order to remove the cell size limitation and to make cellular manufacturing systems more flexible, a manufacturing cell has been equipped with a mobile robot which moves from machine to machine, much like its human counterpart would do. To further acquire a portion of the attributes (intelligence and adaptability of the human worker) lost in robotizing the traditional manned manufacturing cell, a “limited” knowledge-based cell control algorithm is developed. This algorithm maintains standard pull control logic as its underlying basis while utilizing heuristically, knowledge of the ongoing processes and their current status. In this paper, an analysis of this ‘new’ predictive pull-based controller is presented. The simulation results from the predictive pull-based controller are evaluated and compared with the standard pull control logic, first come first serve and shortest distance. A test-bed of an unmanned manufacturing cell with a self-propelled mobile robot was used for experimental verification. Simulation and experimental results show that the degree of improvements gained in the cell performance over the standard pull control method is dependent on the robot’s mobility velocity and how much part inspection/part rejection is performed within the cell. In conclusion, the predictive pull-based algorithm is relatively simple variation of pull control which has the ability of enhancing the “near optimal” pull cell control performance.     

基于Boltzmann 机神经网络认知机制的机器人趋光控制

针对移动机器人未知环境下的趋光控制问题,模拟人或动物“感知-行动”认知机制,对具有趋光特性的移动机器人进行设计,提出一种基于Boltzmann机神经网络的趋光控制方法。该方法首先应用知识集对机器人趋光控制器的Boltzmann机神经网络进行趋光训练;然后应用Boltzmann机神经网络的运行机制实现趋光控制。仿真实验表明,该方法能够提高机器人学习的控制精度。     

一种用于CBR推理机的案例学习算法研究*

案例学习是CBR(Case-Based Reasoning)推理机的重要环节,但由于案例的多样性以及对领域的依赖性,导致CBR系统中案例自动生成困难的问题。针对这一问题,本文提出将seq2seq(Sequence-to-Sequence)模型用于案例学习,通过seq2seq模型自动生成案例,引入attention机制,提高seq2seq模型生成案例的效果,并利用潜在语义分析LSA(Latent Semantic Analysis)对网络爬取语料库进行筛选,利用过滤后的语料库对模型进行训练,提出一种基于三元组的评估方法,对生成案例进行评估和存储,从而实现CBR推理机的自主学习。最后将改进的案例学习系统应用到实际的智能机器人上进行验证,测试结果表明该方法具有可行性,且能够有效提高机器人的智能性及易用性。     

A point and distance constraint based 6R robot calibration method through machine vision

This paper proposes a 6R robot closed-loop kinematic calibration method to improve absolute position accuracy with point and distance constraints though machine vision. In the calibration process, a camera attached to the mounting plate of the robot is used to capture a fixed reference sphere as a point constraint and to record robot joint angles and gauge block lengths that are used as a distance constraint. A first-order difference quotient is used to calculate the Jacobian matrix in the joint parameter identification process. The Staübli TX60 robot is successfully calibrated using the proposed method. After calibration, the average distance error of robot motion is decreased from 2.05 mm to 0.24 mm.