数控车床及测量设备防碰撞系统的研究
|
摘要
数控机床是由NC程序控制加工的,其可由手工编写、软件自动生成等,不可避免含有错误,将会导致刀具损坏或设备事故,因此在实际加工之前都要进行NC程序可靠性的验证。目前数控程序检验方法主要有试切和刀具轨迹仿真。试切法成本高、周期长,不符合经济生产的要求。而刀具轨迹仿真没有考虑机床加工环境,直观性较差。因此三维动态模拟数控机床加工过程是一个重要的研究方向。
在智能坐标测量机的零件位姿识别系统中,为使虚拟零件与实际零件的摆放位姿一致,须建立起精确的零件坐标系,因此需要获得零件在工作台上的放置位置以及放置方向。零件中心位置以及放置方向已经利用零件图像的边缘矩得以解决,详细请见赵金才的博士论文。然而要达到零件的精确定位,仍需要获得零件各个角点在机器坐标系下的坐标。
本文主要工作包括以下几个方面:
1.提出数控车床防碰撞检测系统的总框架,并在VC环境下开发了NC程序检错与纠错模块,对NC程序进行了语法检查和防碰撞检查,经实验验证,系统已经满足基本要求。
2.采用传统Tsai非共面标定算法标定安装在坐标测量机上的摄像机,利用坐标测量机的精确移动和白色陶瓷标准球的良好光学特性,巧妙设计了标定数据采集方案,通过开发数据采集程序以及标定程序使整个标定过程自动化,操作简单、精度高,且只有几分钟就可以完成标定。
3.结合标准Hough算法和最小二乘直线拟合法检测零件角点在图像帧中的坐标,经过摄像机的坐标转换可得到角点在机器坐标系下的坐标;并利用图像帧中的角点坐标反求了零件上表面的Z坐标。算法稳定性好,精度高,通过实验验证满足精度要求。
The numerical control (NC)lathe is controlled by NC program, which is written by hand, or generated by the software automatically. It contains the mistake unavoidably, will cause the damage of the cutter or equipment accident, and must be verified before coming into use. At present, the major methods of verifying NC programs are‘Trial Cut’and‘Cutter Orbit Emulation’.‘Trial Cut’method costs high and has a long cycle, so it doesn’t meet the needs of economic production.‘Cutter Orbit Emulation’method doesn’t consider machines condition. Therefore the three-dimension dynamic simulation of NC lathe processing is an important research direction.
In the recognition system of part location of intellectual coordinate measurement machine(CMM), in order to make the virtual part’s position accord with the real one’s, the position and direction of the real part must be gotten to set up an accurate coordinate system. To get the center point and direction of a part, please refer to Zhao Jincai's Doctor Thesis. This article focuses on obtaining the coordinate of angular points for accurate part localization.
The following work has being done in this thesis:
1. Put forward the whole frame of the anti-collision system of the NC lathe, and develop error checking and correcting module to do grammar checking and colliding detecting, it is proved by experiments to be able to cover the basic requirements.
2. Adopt traditional Tsai non-coplanar calibration algorithm to calibrate camera set on the CMM. The scheme of gathering calibration data is designed utilizing accurate movement of the CMM and the good optics characteristic of white standard ball. By developing data collecting and calibrating programs, the whole process is made automated, simple and precise, which can be done within few minutes.
3. Detect angular points in the parts’images with standard Hough algorithm and least square method,and calculate the Z coordinate of upper surface of part with the angular points .The algorithm has high accuracy and stability.
在智能坐标测量机的零件位姿识别系统中,为使虚拟零件与实际零件的摆放位姿一致,须建立起精确的零件坐标系,因此需要获得零件在工作台上的放置位置以及放置方向。零件中心位置以及放置方向已经利用零件图像的边缘矩得以解决,详细请见赵金才的博士论文。然而要达到零件的精确定位,仍需要获得零件各个角点在机器坐标系下的坐标。
本文主要工作包括以下几个方面:
1.提出数控车床防碰撞检测系统的总框架,并在VC环境下开发了NC程序检错与纠错模块,对NC程序进行了语法检查和防碰撞检查,经实验验证,系统已经满足基本要求。
2.采用传统Tsai非共面标定算法标定安装在坐标测量机上的摄像机,利用坐标测量机的精确移动和白色陶瓷标准球的良好光学特性,巧妙设计了标定数据采集方案,通过开发数据采集程序以及标定程序使整个标定过程自动化,操作简单、精度高,且只有几分钟就可以完成标定。
3.结合标准Hough算法和最小二乘直线拟合法检测零件角点在图像帧中的坐标,经过摄像机的坐标转换可得到角点在机器坐标系下的坐标;并利用图像帧中的角点坐标反求了零件上表面的Z坐标。算法稳定性好,精度高,通过实验验证满足精度要求。
The numerical control (NC)lathe is controlled by NC program, which is written by hand, or generated by the software automatically. It contains the mistake unavoidably, will cause the damage of the cutter or equipment accident, and must be verified before coming into use. At present, the major methods of verifying NC programs are‘Trial Cut’and‘Cutter Orbit Emulation’.‘Trial Cut’method costs high and has a long cycle, so it doesn’t meet the needs of economic production.‘Cutter Orbit Emulation’method doesn’t consider machines condition. Therefore the three-dimension dynamic simulation of NC lathe processing is an important research direction.
In the recognition system of part location of intellectual coordinate measurement machine(CMM), in order to make the virtual part’s position accord with the real one’s, the position and direction of the real part must be gotten to set up an accurate coordinate system. To get the center point and direction of a part, please refer to Zhao Jincai's Doctor Thesis. This article focuses on obtaining the coordinate of angular points for accurate part localization.
The following work has being done in this thesis:
1. Put forward the whole frame of the anti-collision system of the NC lathe, and develop error checking and correcting module to do grammar checking and colliding detecting, it is proved by experiments to be able to cover the basic requirements.
2. Adopt traditional Tsai non-coplanar calibration algorithm to calibrate camera set on the CMM. The scheme of gathering calibration data is designed utilizing accurate movement of the CMM and the good optics characteristic of white standard ball. By developing data collecting and calibrating programs, the whole process is made automated, simple and precise, which can be done within few minutes.
3. Detect angular points in the parts’images with standard Hough algorithm and least square method,and calculate the Z coordinate of upper surface of part with the angular points .The algorithm has high accuracy and stability.
引文
[1]王锐,STEP-NC 数控加工的仿真技术研究,硕士学位论文,山东大学,2004
[2]赵金才,坐标测量系统零件信息提取与位姿自动识别的研究,博士学位论文,天津大学,2005
[3]Xiao Tianyuan, Han Xiangli Yang Gang,Virtual machining environment-GMPS,IEEE Symposium on Emerging Technologies & Factory Automation, 1996,167～171
[4]王洪刚,虚拟数控技术仿真系统的研究,硕士学位论文,浙江大学,2004
[5]杨慕升,王爱民,虚拟环境下 CAD/CAM 及产品开发技术,山东工程学院学报,2002,16(4),1～5
[6]罗亚波,陈立方,肖田元,远程数控仿真的一种编程技术,武汉理工大学学报,2004,28(1),62～65
[7]陈淘,叶佩青,汪劲松,数控加工过程的计算机仿真技术研究,计算机应用研究,2004,21(9),194～195、198
[8]彼得·斯密德,数控编程手册(罗学克,刘瑛等译),北京:化学工业出版社,2005,48～522
[9]国家机械工业局,JB/T 3208—1999,数控机床穿孔带程序段格式中的准备功能 G 和辅助功能 M 的代码,北京:机械科学研究院,1999-01-01
[10]顾京,数控机床加工程序编制,北京:机械工业出版社,2005
[11]王卫兵,数控编程 100 例,北京:机械工业出版社,2004
[12]南雁,基于虚拟数控加工的 NC 代码翻译,硕士学位论文,西北工业大学,2003
[13]徐晰,刘书桂,CMM/CAD 集成系统中 IGES 后处理器的开发,天津大学学报,2003,36(1),38～42
[14]徐士良,计算机常用算法,北京:清华大学出版社,1995
[15]马新辉,智能三坐标测量机的研究,天津大学博士学位论文,2002
[16]马新辉,张国雄,王建利,刘书桂,智能三坐标测量机中零件位置自动识别系统,仪器仪表学报,2003,24(2):165～170
[17]罗飞路,傅恩锡,CCD 摄像机内外部参数快速准确的标定方法,湖南大学学报,1997,24(2),71～74
[18]皱凤娇,苏显渝等,基于共面点的摄像机线性标定法,光电工程,2005,32(4),70～74
[19]Tsai R Y, A versatile camera calibration technique for high accuracy 3D machine vision metrology using of the shelf TV camera and lenses, IEEE Journal of Automation, 1987, 3(4): 323～344
[20]Lenz R K, Tsai R Y, Techniques for calibration of the scale factor and image center for high accuracy 3D machine vision metrology, IEEE Trans. on PAMI, 1988, 10(5): 713～720
[21]孙即祥,模式识别中的特征提取与计算机视觉不变量,北京:国防工业出版社,2001,148～159
[22]陈利军,图像角点检测和匹配算法的研究,硕士学位论文,西安电子科技大惠思等,基于上下文敏感度的动态图像目标识别,西安交通大学K C,et al,A Servey of Thresholding Techniques, from Gray-Level Histogram,IEEE Trans on M. Stephens, A combined corner and edge detector, Fourth Alvey 测绘信息与detect lines and 振江,王渝等,采用 Hough 变化和灰度变化的图像角点检测法,北京理工学,2005
[23]姚磊,基于 Hough 变换和不变距的图像模式识别技术研究,硕士学位论文,燕山大学,2006
[24]王建勇,周晓光等,一种改进的直线检测算法,计算机工程,2006,32(16),172～173、187
[25]贾晓琳,吴学报,2005,39(6),590～593
[26]楚锋,基于计算机视觉的三维测量实现方法研究,硕士学位论文,东北大学,2006
[27]何斌,马天予等,Visual C++数字图像处理,北京:人民邮电出版社,2001
[28]Sahoo PK,Soltanti S,Wong AComputer Vision Graphics Image Process, 1998,41: 233～260
[29]Otsu,A Threshold Selection MethodSMC-9, 1979: 62～66
[30]C. Harris andVision Conference, 1988,.147～151
[31]谢东海,詹总谦等,改进 Harris 算子用于点特征的精确定位,工程,2003,28(2),22～23
[32]DUDA RO,HART PE,Use of the Hough transformation tocurves in pictures,Commun ACM, 1972, 15(1), 11～15
[33]蔡大学学报,2005,25(9),796～799
[2]赵金才,坐标测量系统零件信息提取与位姿自动识别的研究,博士学位论文,天津大学,2005
[3]Xiao Tianyuan, Han Xiangli Yang Gang,Virtual machining environment-GMPS,IEEE Symposium on Emerging Technologies & Factory Automation, 1996,167～171
[4]王洪刚,虚拟数控技术仿真系统的研究,硕士学位论文,浙江大学,2004
[5]杨慕升,王爱民,虚拟环境下 CAD/CAM 及产品开发技术,山东工程学院学报,2002,16(4),1～5
[6]罗亚波,陈立方,肖田元,远程数控仿真的一种编程技术,武汉理工大学学报,2004,28(1),62～65
[7]陈淘,叶佩青,汪劲松,数控加工过程的计算机仿真技术研究,计算机应用研究,2004,21(9),194～195、198
[8]彼得·斯密德,数控编程手册(罗学克,刘瑛等译),北京:化学工业出版社,2005,48～522
[9]国家机械工业局,JB/T 3208—1999,数控机床穿孔带程序段格式中的准备功能 G 和辅助功能 M 的代码,北京:机械科学研究院,1999-01-01
[10]顾京,数控机床加工程序编制,北京:机械工业出版社,2005
[11]王卫兵,数控编程 100 例,北京:机械工业出版社,2004
[12]南雁,基于虚拟数控加工的 NC 代码翻译,硕士学位论文,西北工业大学,2003
[13]徐晰,刘书桂,CMM/CAD 集成系统中 IGES 后处理器的开发,天津大学学报,2003,36(1),38～42
[14]徐士良,计算机常用算法,北京:清华大学出版社,1995
[15]马新辉,智能三坐标测量机的研究,天津大学博士学位论文,2002
[16]马新辉,张国雄,王建利,刘书桂,智能三坐标测量机中零件位置自动识别系统,仪器仪表学报,2003,24(2):165～170
[17]罗飞路,傅恩锡,CCD 摄像机内外部参数快速准确的标定方法,湖南大学学报,1997,24(2),71～74
[18]皱凤娇,苏显渝等,基于共面点的摄像机线性标定法,光电工程,2005,32(4),70～74
[19]Tsai R Y, A versatile camera calibration technique for high accuracy 3D machine vision metrology using of the shelf TV camera and lenses, IEEE Journal of Automation, 1987, 3(4): 323～344
[20]Lenz R K, Tsai R Y, Techniques for calibration of the scale factor and image center for high accuracy 3D machine vision metrology, IEEE Trans. on PAMI, 1988, 10(5): 713～720
[21]孙即祥,模式识别中的特征提取与计算机视觉不变量,北京:国防工业出版社,2001,148～159
[22]陈利军,图像角点检测和匹配算法的研究,硕士学位论文,西安电子科技大惠思等,基于上下文敏感度的动态图像目标识别,西安交通大学K C,et al,A Servey of Thresholding Techniques, from Gray-Level Histogram,IEEE Trans on M. Stephens, A combined corner and edge detector, Fourth Alvey 测绘信息与detect lines and 振江,王渝等,采用 Hough 变化和灰度变化的图像角点检测法,北京理工学,2005
[23]姚磊,基于 Hough 变换和不变距的图像模式识别技术研究,硕士学位论文,燕山大学,2006
[24]王建勇,周晓光等,一种改进的直线检测算法,计算机工程,2006,32(16),172～173、187
[25]贾晓琳,吴学报,2005,39(6),590～593
[26]楚锋,基于计算机视觉的三维测量实现方法研究,硕士学位论文,东北大学,2006
[27]何斌,马天予等,Visual C++数字图像处理,北京:人民邮电出版社,2001
[28]Sahoo PK,Soltanti S,Wong AComputer Vision Graphics Image Process, 1998,41: 233～260
[29]Otsu,A Threshold Selection MethodSMC-9, 1979: 62～66
[30]C. Harris andVision Conference, 1988,.147～151
[31]谢东海,詹总谦等,改进 Harris 算子用于点特征的精确定位,工程,2003,28(2),22～23
[32]DUDA RO,HART PE,Use of the Hough transformation tocurves in pictures,Commun ACM, 1972, 15(1), 11～15
[33]蔡大学学报,2005,25(9),796～799