面向复杂形体表面操作的工业机器人路径规划与仿真研究

发布时间：2018-10-24 10:44

【摘要】：随着社会用工成本不断上升,工业机器人的应用越来越广泛。在工业机器人的生产作业中,其面向的被加工对象也越来越复杂。因而,研究复杂形体表面的机器人作业方法,对提升我国工业机器人加工技术具有重要意义。本文针对复杂形体表面作业,提出了工业机器人路径规划与仿真方法。首先,对给定的复杂形体,利用人机交互勾绘和特征提取技术,并结合位姿数据转化和属性操作,快速完成工业机器人任务路径的设计。其次,通过构建模型的自适应层次包围盒,并根据机械臂结构和非接触角分析,实现工业机器人的快速碰撞检测。再次,利用遗传算法规划形体表面路径加工的序列,并通过碰撞检测模型筛选出非碰撞作业路径。最后,通过工业机器人仿真系统验证了本文研究成果的可行性。本文主要研究内容如下:研究了复杂形体表面机器人作业任务路径快速构建技术。通过形体表面交互和点间切割操作,快速构建与形体表面紧密贴合的几何路径;针对难以精确勾绘的模型边界,通过几何和拓扑关系提取特征边,并对于特征边序列化,从而完成形体边界的几何路径快速构建;对构建的几何路径进行机器人任务路径的数据转化、变换和补偿等操作,并设置任务路径关键工艺参数,进而完成机器人任务路径的快速构建;最后,通过仿真实验验证了方法的可行性。研究了复杂形体表面路径机器人作业仿真碰撞检测技术。针对工业机器人作业仿真碰撞检测问题,本文避开了传统构建繁杂的层次包围盒方法,通过复杂形体表面路径机器人作业仿真,探究了基于AOAAE复杂形体表面路径机器人作业的碰撞检测,加快了碰撞检测的速度;本文基于机械臂的特征分析,利用大臂运动范围的广泛性,提出机械臂非接触角的概念,并利用机械臂的结构与非接触角,分析了机械臂自相交碰撞检测,改善了碰撞检测的实时性。研究了基于遗传算法的形体表面复杂路径规划技术。本文以映射形体表面点阵作为优化目标点,通过对目标点初始参数和遗传算子设计,完成形体表面点阵序列规划;建立了形体表面曲线的简化模型,并结合贪心法思想,确定相邻路径的连接点的位置,给出了不规则形体表面多路径加工序列的方法;在上述方法的基础上,进一步做无碰撞路径筛选操作,并获得到无碰撞近似最优路径。开发了工业机器人的复杂形体表面路径规划与仿真系统。基于本文研究内容和方法,开发了工业机器人的复杂形体表面路径规划与仿真系统,其中设计了离线软件的功能模块和UI界面,并开发了虚拟示教器和路径规划器等工具。在仿真应用实例中,搭建了机器人作业环境,构建了任务路径,利用机器人刻字、焊接仿真作业,综合验证了本文所提方法的可行性。
[Abstract]:With the increasing of social labor costs, industrial robots are more and more widely used. In the production of industrial robots, the objects they face are becoming more and more complex. Therefore, it is of great significance to study the working methods of the robot on the surface of complex objects in order to improve the machining technology of industrial robots in China. In this paper, a path planning and simulation method of industrial robot is proposed for complex object surface operation. Firstly, the task path design of industrial robot is accomplished quickly by means of human-computer interactive mapping and feature extraction techniques, combined with the transformation of position and pose data and attribute operation for a given complex object. Secondly, by constructing adaptive hierarchical bounding box and analyzing the structure of manipulator and non-contact angle, the fast collision detection of industrial robot is realized. Thirdly, genetic algorithm is used to program the sequence of surface path processing, and the non-collision operation path is screened by collision detection model. Finally, the feasibility of the research results is verified by the industrial robot simulation system. The main contents of this paper are as follows: the fast task path construction technology of complex surface robot is studied. Through the interaction of the surface of the body and the operation of cutting between points, the geometric path closely aligned with the surface of the body is constructed quickly, and for the model boundary which is difficult to draw accurately, the feature edges are extracted by geometric and topological relations, and the feature edges are serialized. Thus the geometry path of body boundary is constructed quickly, the data transformation, transformation and compensation of the robot task path are carried out, and the key process parameters of the task path are set up. Finally, the feasibility of the method is verified by simulation experiments. The simulation collision detection technology of complex surface path robot is studied in this paper. Aiming at the problem of industrial robot operation simulation collision detection, this paper avoids the traditional method of constructing complicated hierarchical bounding box, and simulates the operation of robot with complex surface path. This paper probes into the collision detection of the robot with complex surface path based on AOAAE, which accelerates the speed of collision detection. Based on the characteristic analysis of the manipulator, the concept of the non-contact angle of the manipulator is put forward based on the extensiveness of the moving range of the arm. Using the structure and non-contact angle of the manipulator, the self-intersecting collision detection of the manipulator is analyzed, which improves the real-time performance of the collision detection. The complex path planning technology of shape surface based on genetic algorithm is studied. In this paper, we take the mapping surface lattice as the optimization target point, design the initial parameters of the target point and genetic operator to complete the planning of the surface lattice sequence of the object surface, establish a simplified model of the surface curve of the body, and combine the greedy method with the idea of greedy method. The location of the joining points of the adjacent paths is determined, and the method of multipath processing sequence on the surface of irregular bodies is given. On the basis of the above methods, the collision free path selection operation is further done, and the collision free approximate optimal path is obtained. A complex surface path planning and simulation system for industrial robot is developed. Based on the contents and methods of this paper, a complex object surface path planning and simulation system of industrial robot is developed, in which the function module and UI interface of off-line software are designed, and the tools such as virtual teacher and path planner are developed. In the simulation application example, the robot working environment is built, the task path is constructed, and the feasibility of the proposed method is verified synthetically by using the robot engraving and welding simulation operation.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242.2

【参考文献】