面向机器人离线编程的路径与轨迹规划算法研究

发布时间：2018-04-28 22:05

本文选题：离线编程 + STL模型　；参考：《广东工业大学》2017年硕士论文

【摘要】：机器人编程方法可分为在线示教编程与离线编程,常用的在线示教编程方法具有精度低、示教繁琐等缺点。离线编程方法能克服在线示教编程的缺点,从而提高生产效率与保证加工精度。离线编程主要采用从工件的三维CAD模型直接提取机器人运动轨迹的方法完成机器人对工件的表面加工。本文深入研究面向机器人离线编程系统的关键技术——轨迹规划。主要研究内容如下:1、首先,本文研究了一种基于平衡二叉树动态搜索的STL模型生成路径的算法。STL格式文件包含着待加工曲面的表面信息,可作为获取路径的来源文件。通过三维CAD建模获取包含待加工曲面的STL模型然后通过多个平行切平面与STL模型的三角面片相交,并通过切片平面与三角面片的求交运算得到路径的数据点组。运用平衡二叉树动态搜索算法对这些排列杂乱无章的数据点组进行优化排序后形成一条完整的路径线。2、其次,本文采用3次NURBS曲线的反求算法对路径进行优化。按照STL模型的切片算法得到的路径,是由分布不规则的离散点连接而成的。为了使路径更加连贯,采用3次NURBS曲线对原路径进行优化。提出自适应的参数化法算法确定节点矢量,在已知路径型值点的情况下,求得能够准确描述路径的3次NURBS曲线的节点矢量和控制点。3、再次,本文研究基于弦截迭代法的机器人轨迹规划。在已知能够准确描述路径的3次NURBS曲线的表达式的基础上,根据弓高误差预测出插补点处的节点参数,使用弦截迭代法对预测的节点参数进行迭代求出真实的节点参数,保证理想速度与实际速度之间的波动率在一定范围内。4、最后,运用MFC与Open GL技术开发了轨迹生成系统。根据本文研究的路径生成、路径优化与轨迹规划的算法,提出系统的体系结构、系统界面、模块功能,对提出的机器人运动轨迹生成方法进行验证。在系统中导入猴型雕刻工件的STL模型生成机器人运动轨迹并进行实际加工,并且与仅基于STL切片算法得到的运动轨迹进行对比验证本文研究算法的有效性。
[Abstract]:Robot programming methods can be divided into online teaching programming and off-line programming. The common online teaching programming methods have the disadvantages of low precision and tedious teaching. Off-line programming method can overcome the shortcomings of online teaching programming, thus improve production efficiency and ensure machining accuracy. Off-line programming mainly uses 3D CAD model of workpiece to extract robot motion trajectory directly to finish the surface machining of workpiece. In this paper, trajectory planning, the key technology of robot off-line programming system, is studied. The main contents of this paper are as follows: 1. Firstly, this paper studies an algorithm of path generation for STL model based on balanced binary tree dynamic search. STL format file contains the surface information of the surface to be processed and can be used as the source file to obtain the path. The STL model containing the surface to be machined is obtained by 3D CAD modeling, then the triangulated surface of the STL model is intersected by several parallel tangent planes, and the data set of the path is obtained by the intersection of the slicing plane and the triangulated surface. The balanced binary tree dynamic search algorithm is used to optimize and sort these disordered data points to form a complete path line. Secondly, this paper uses the inverse algorithm of three NURBS curves to optimize the path. According to the slicing algorithm of STL model, the path is connected by scattered points with irregular distribution. In order to make the path more coherent, the original path is optimized by using three NURBS curves. An adaptive parameterization algorithm is proposed to determine the node vector. When the path type value points are known, the node vector and control point. 3, which can accurately describe the cubic NURBS curve of the path, are obtained. In this paper, the trajectory planning of robot based on chord-truncated iteration method is studied. Based on the known expression of the cubic NURBS curve which can accurately describe the path, the node parameters at the interpolation point are predicted according to the error of the bow height, and the real node parameters are obtained by iterating the predicted node parameters by using the chord section iteration method. The fluctuation rate between ideal velocity and actual velocity is guaranteed to be within a certain range. Finally, a trajectory generation system is developed by using MFC and Open GL techniques. According to the algorithms of path generation, path optimization and trajectory planning studied in this paper, the system architecture, system interface and module functions are proposed, and the proposed robot motion trajectory generation method is verified. The STL model of monkey carving was imported into the system to generate the robot motion trajectory and processed in practice, and compared with the motion trajectory obtained only based on the STL slicing algorithm, the validity of the proposed algorithm was verified.
【学位授予单位】：广东工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】