六自由度工业机器人运动学参数辨识方法研究
发布时间：2019-06-14 20:28
【摘要】：随着工业机器人的广泛应用,作为衡量其性能指标之一的末端位置精度,已经逐渐引起人们的关注。由于各种误差因素的影响,工业机器人末端实际位置与理论位置总是存在着一定的误差,这严重影响了工业机器人在高精度要求场合的应用与推广。目前,标定技术是提高工业机器人末端位置精度的主要方法。本文围绕六自由度工业机器人的标定方法和误差补偿问题,以某型号工业机器人为对象,主要开展了以下几方面的工作:针对工业机器人标定问题,结合工业机器人本体结构特点,利用DH(Denavit-Hartenberg)方法建立运动学模型,推导出末端位置与DH参数的关系,并通过仿真与实验验证了所建立的运动学模型。推导出DH参数误差与末端位置误差的转换关系,进一步得到了工业机器人运动学误差模型。对工业机器人运动学参数的线性相关性进行了分析,得到了线性相关的参数以及其对运动学参数辨识结果的影响规律。根据运动学模型与误差模型,提出利用最小二乘法对六自由度工业机器人的运动学参数进行求解。为了简化运动学参数辨识步骤,本文提出一种遗传禁忌搜索算法,此算法不需要对误差模型进行分析与变换,而将参数误差求解看作是一个优化问题,利用遗传算法与禁忌搜索算法相结合的方法进行最优值搜索,最后获得最优适应度函数值。基于Robotics等工具箱与MATLAB的GUI界面,编制了参数辨识软件,分别利用最小二乘法和遗传禁忌搜索算法对工业机器人的运动学参数进行仿真辨识,确定了参数辨识算法需要的最少数据量。仿真结果显示,在进行参数辨识和补偿后,末端绝对位置精度得到明显提升,基于最小二乘法的参数辨识后,机器人末端最大方向误差从3mm,下降到0.005mm。基于遗传禁忌搜索算法参数辨识后,末端最大方向误差下降到0.008mm。以关节臂式坐标测量机作为测量工具进行了相关实验研究。首先测量了工业机器人末端在测量机坐标系下的空间位置坐标;然后再利用算法将空间位置坐标转换到工业机器人坐标系中,并分别利用上述两种辨识方法进行运动学参数辨识。一般情况下,机器人末端绝对位置精度通常要求在5mm以内,辨识后,最大方向误差从15mm下降到0.7mm和1.4mm以内,可以满足机器人在一般情况下的使用要求;最后将两组实验结果进行对比分析,基于最小二乘法的参数辨识效果较好,但是基于遗传禁忌搜索算法的参数辨识算法的辨识步骤较简单,效率更高,可以专注于算法的优化。
[Abstract]:With the wide application of industrial robots, as one of the performance indicators of industrial robots, the end position accuracy has gradually attracted people's attention. Due to the influence of various error factors, there are always some errors between the actual position and the theoretical position of the industrial robot, which seriously affects the application and popularization of the industrial robot in the case of high precision requirements. At present, calibration technology is the main method to improve the terminal position accuracy of industrial robots. In this paper, focusing on the calibration method and error compensation of six-degree-of-freedom industrial robot, taking a certain type of industrial robot as the object, the following work is carried out: aiming at the calibration problem of industrial robot, combined with the structural characteristics of industrial robot body, the kinematic model is established by DH (Denavit-Hartenberg) method, and the relationship between end position and DH parameters is derived. The kinematic model is verified by simulation and experiment. The conversion relationship between DH parameter error and end position error is derived, and the kinematic error model of industrial robot is further obtained. The linear correlation of kinematic parameters of industrial robot is analyzed, and the linear correlation parameters and their influence on the identification results of kinematic parameters are obtained. According to the kinematic model and error model, the least square method is proposed to solve the kinematic parameters of six-degree-of-freedom industrial robot. In order to simplify the steps of kinematic parameter identification, a genetic Tabu search algorithm is proposed in this paper, which does not need to analyze and transform the error model, but regards the parameter error solution as an optimization problem. The optimal value search is carried out by using the combination of genetic algorithm and Tabu search algorithm, and finally the optimal fitness function value is obtained. Based on the GUI interface between Robotics and MATLAB, the parameter identification software is compiled. The kinematic parameters of industrial robot are simulated and identified by using the least square method and genetic Tabu search algorithm, and the minimum amount of data required by the parameter identification algorithm is determined. The simulation results show that the absolute position accuracy of the end is improved obviously after parameter identification and compensation. After the parameter identification based on the least square method, the maximum direction error at the end of the robot decreases from 3mm to 0.005mm. After parameter identification based on genetic Tabu search algorithm, the maximum direction error at the end is reduced to 0.008mm. The joint arm coordinate measuring machine is used as the measuring tool to carry on the related experimental research. Firstly, the spatial position coordinates of the industrial robot end in the measuring machine coordinate system are measured, and then the spatial position coordinates are transformed into the industrial robot coordinate system by using the algorithm, and the kinematic parameters are identified by using the above two identification methods respectively. In general, the absolute position accuracy of the end of the robot is usually within 5mm. After identification, the maximum direction error decreases from 15mm to 0.7mm and 1.4mm, which can meet the requirements of the robot in general. Finally, the two groups of experimental results are compared and analyzed. The parameter identification effect based on least square method is better, but the identification step of parameter identification algorithm based on genetic Tabu search algorithm is simple, the efficiency is higher, and the algorithm can focus on the optimization of the algorithm.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242.2
本文编号：2499668
[Abstract]:With the wide application of industrial robots, as one of the performance indicators of industrial robots, the end position accuracy has gradually attracted people's attention. Due to the influence of various error factors, there are always some errors between the actual position and the theoretical position of the industrial robot, which seriously affects the application and popularization of the industrial robot in the case of high precision requirements. At present, calibration technology is the main method to improve the terminal position accuracy of industrial robots. In this paper, focusing on the calibration method and error compensation of six-degree-of-freedom industrial robot, taking a certain type of industrial robot as the object, the following work is carried out: aiming at the calibration problem of industrial robot, combined with the structural characteristics of industrial robot body, the kinematic model is established by DH (Denavit-Hartenberg) method, and the relationship between end position and DH parameters is derived. The kinematic model is verified by simulation and experiment. The conversion relationship between DH parameter error and end position error is derived, and the kinematic error model of industrial robot is further obtained. The linear correlation of kinematic parameters of industrial robot is analyzed, and the linear correlation parameters and their influence on the identification results of kinematic parameters are obtained. According to the kinematic model and error model, the least square method is proposed to solve the kinematic parameters of six-degree-of-freedom industrial robot. In order to simplify the steps of kinematic parameter identification, a genetic Tabu search algorithm is proposed in this paper, which does not need to analyze and transform the error model, but regards the parameter error solution as an optimization problem. The optimal value search is carried out by using the combination of genetic algorithm and Tabu search algorithm, and finally the optimal fitness function value is obtained. Based on the GUI interface between Robotics and MATLAB, the parameter identification software is compiled. The kinematic parameters of industrial robot are simulated and identified by using the least square method and genetic Tabu search algorithm, and the minimum amount of data required by the parameter identification algorithm is determined. The simulation results show that the absolute position accuracy of the end is improved obviously after parameter identification and compensation. After the parameter identification based on the least square method, the maximum direction error at the end of the robot decreases from 3mm to 0.005mm. After parameter identification based on genetic Tabu search algorithm, the maximum direction error at the end is reduced to 0.008mm. The joint arm coordinate measuring machine is used as the measuring tool to carry on the related experimental research. Firstly, the spatial position coordinates of the industrial robot end in the measuring machine coordinate system are measured, and then the spatial position coordinates are transformed into the industrial robot coordinate system by using the algorithm, and the kinematic parameters are identified by using the above two identification methods respectively. In general, the absolute position accuracy of the end of the robot is usually within 5mm. After identification, the maximum direction error decreases from 15mm to 0.7mm and 1.4mm, which can meet the requirements of the robot in general. Finally, the two groups of experimental results are compared and analyzed. The parameter identification effect based on least square method is better, but the identification step of parameter identification algorithm based on genetic Tabu search algorithm is simple, the efficiency is higher, and the algorithm can focus on the optimization of the algorithm.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242.2
【参考文献】
相关期刊论文 前10条
1 熊杰;杨东升;王允森;;遗传禁忌搜索算法在工业机器人结构参数辨识上的应用[J];组合机床与自动化加工技术;2015年12期
2 齐飞;平雪良;刘洁;蒋毅;;工业机器人参数辨识及误差补偿方法研究[J];机械传动;2015年09期
3 白云飞;丛明;杨小磊;刘冬;;基于6参数模型的6R串联机器人运动学参数辨识[J];机器人;2015年04期
4 李睿;曲兴华;;工业机器人运动学参数标定误差不确定度研究[J];仪器仪表学报;2014年10期
5 王晓强;王帅军;刘建亭;;基于MATLAB的IRB2400工业机器人运动学分析[J];机床与液压;2014年03期
6 蔡锦达;张剑皓;秦绪祥;;六轴工业机器人的参数辨识方法[J];控制工程;2013年05期
7 谭民;王硕;;机器人技术研究进展[J];自动化学报;2013年07期
8 王智兴;樊文欣;张保成;史源源;;基于Matlab的工业机器人运动学分析与仿真[J];机电工程;2012年01期
9 张博;魏振忠;张广军;;机器人坐标系与激光跟踪仪坐标系的快速转换方法[J];仪器仪表学报;2010年09期
10 王文;林铿;高贯斌;陈子辰;;关节臂式坐标测量机角度传感器偏心参数辨识[J];光学精密工程;2010年01期
相关博士学位论文 前3条
1 张晓平;六自由度关节型机器人参数标定方法与实验研究[D];华中科技大学;2013年
2 高贯斌;关节臂式坐标测量机自标定方法与误差补偿研究[D];浙江大学;2010年
3 张永贵;喷漆机器人若干关键技术研究[D];西安理工大学;2008年
相关硕士学位论文 前10条
1 张虎;面向标定的工业机器人建模及参数辨识方法研究[D];哈尔滨工业大学;2015年
2 熊杰;六关节机器人误差补偿技术研究与实现[D];中国科学院研究生院（沈阳计算技术研究所）;2015年
3 时定兵;基于点约束的机器人运动学参数标定技术研究[D];南京理工大学;2014年
4 张越;机器人运动学参数辨识及冗余参数研究[D];哈尔滨工业大学;2013年
5 侯士杰;工业机器人结构参数辨识与位姿误差补偿研究[D];南京航空航天大学;2012年
6 龚星如;六自由度工业机器人运动学标定的研究[D];南京航空航天大学;2012年
7 夏天;工业机器人运动学标定及误差分析研究[D];上海交通大学;2009年
8 刘建华;六自由度串联机器人运动仿真研究[D];燕山大学;2008年
9 南小海;6R型工业机器人标定算法与实验研究[D];华中科技大学;2008年
10 王斌;关节臂式三坐标测量系统数学模型及标定技术的研究[D];天津大学;2007年
，本文编号：2499668
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2499668.html
