面向复杂自由曲面的喷涂机器人作业规划方法研究与实现

发布时间：2018-06-15 00:09

本文选题：复杂曲面喷涂 + 曲面分片　；参考：《东南大学》2016年硕士论文

【摘要】：本论文课题以“高档数控机床与基础制造设备”国家重大科技专项“15kg喷涂机器人成套装备”项目为背景,研究并实现面向复杂曲面的喷涂机器人作业规划方法。由于复杂自由曲面表面特征不同于一般的平面与拉伸曲面,本文研究并实现一种新的方法来解决复杂曲面上的喷涂机器人作业规划问题,该方法分为三个部分：曲面分片、子片喷涂路径规划和子片连接。复杂自由曲面几何特征(形状、曲率等)和拓扑特征(孔洞等)复杂,这些都增加了喷涂路径规划的难度。因此本文首先分析曲面上的拓扑特征和几何特征,然后对曲面分片,将复杂自由曲面分割为若干具有简单拓扑特征与几何特征的简单子曲面。对于曲面上有孔洞等情形,采用投影法,对曲面进行拓扑分片,把曲面分成几个独立的子片,每一片上不包含孔洞；对于包含圆孔这样的特殊工件,本文通过选取两个特殊平面将曲面分割成几个独立的子片,进行拓扑分片。其次每一分片上可能曲率变化大,通过设置相邻角阈值和最大背离角,对其进行几何分片,使得每一分片上的曲率接近,变化不大。曲面分片之后,通常每一子片并不是平面,其曲率在各方向上都可能变化,为此,针对复杂曲面论文研究并实现了一种基于种子曲线、变速变间距的喷涂路径规划方法以进一步提高喷涂效果。首先确定曲面上一条曲线作为种子曲线,以该种子曲线作为基准线,平移得到曲面上所有喷涂行程。种子曲线选择的原则是,规划好之后曲面上的行程尽量少,而且行程之间不会相交。然后,将每一行程分段,每一段上速度不同,通过对分段速度优化,使得行程方向涂层累积速率近似不变。再对行程间距优化,考虑到多条行程对同一点膜厚会产生影响,这里同时对4条行程优化,提高喷涂规划准确性。由于速度优化与间距优化,都是针对多变量的,采用BOBYQA优化方法,得到所有行程间距以及每一条行程上的分段速度,进而得到子片上的喷枪路径。在子片喷枪路径规划的基础上,为了获得完整的喷枪路径,需要将这些独立的子片路径连接起来,经过所有子片上的喷枪路径,并且行进的路径最短。本文将子片路径连接问题建模为广义TSP问题模型,采用蚁群算法来得到优化解,进而连接子片喷枪路径间的起点与终点,形成一条完整的喷枪路径,并生成中间语言程序。蚁群算法与之前采用的启发式搜索算法如遗传算法相比,收敛速度更快,而且排序优化结果更好即连接路径的长度更短,这些优势在子片喷涂路径数量增加时更明显。最后针对典型工件,仿真验证喷涂路径规划与参数优化的喷涂结果。本文完成了面向复杂自由曲面的喷涂机器人作业规划方法研究与实现,该方法能够针对复杂自由曲面,规划出喷枪路径,经喷涂作业规划系统仿真验证,能较好地满足项目需求。
[Abstract]:Based on the project of "High-grade NC Machine tool and basic Manufacturing equipment" and "15kg spray Robot complete equipment", the paper studies and realizes the operation planning method of spray-robot facing complex curved surface. Because the surface features of complex free-form surfaces are different from those of general plane and stretch surfaces, this paper studies and implements a new method to solve the problem of job planning of spray-robot on complex surfaces. The method is divided into three parts: surface segmentation. Subblade spray path planning and sub-strip connection. The geometric features (shape, curvature, etc.) of complex free-form surfaces and topological features (holes, etc.) are complex, all of which increase the difficulty of spray path planning. In this paper, we first analyze the topological and geometric features of surfaces, and then divide the complex free-form surfaces into several simple subsurfaces with simple topological and geometric features. In the case of a hole in a curved surface, the surface is topologically divided into several separate subslices, each of which does not contain holes, and for a special workpiece such as a circular hole, In this paper, two special planes are selected to divide the surface into several independent subslices. Secondly, the possible curvature changes greatly on each slice. By setting the threshold of adjacent angle and the maximum deviation angle, the geometric partition is carried out to make the curvature of each slice close, and the change is not significant. After the surface is partitioned, each subslice is usually not a plane, and its curvature may change in every direction. Therefore, a seed curve based on the complex surface is studied and implemented in this paper. The spray path planning method with variable speed and variable spacing is used to further improve the spraying effect. First, a curve on the surface is determined as the seed curve, and the seed curve is used as the datum line, and all spray stroke on the surface is obtained by translation. The principle of seed curve selection is that there are as few trips on the surface as possible after planning, and there is no intersecting between them. Then, the speed of each stroke segment and each section is different, and the cumulative rate of the coating is approximately unchanged by optimizing the segmented velocity. Considering the influence of multiple stroke on the thickness of film at the same point, four strokes are optimized at the same time to improve the accuracy of spray planning. Because velocity optimization and spacing optimization are all aimed at multivariable, using BOBYQA optimization method, we can get all travel distance and the segmented velocity on each stroke, and then get the spray gun path on the sub-chip. On the basis of the path planning of the sub-piece spray gun, in order to obtain the complete spray gun path, these independent sub-slice paths need to be connected together, passing through all the spray gun paths on the sub-slices, and the path is the shortest. In this paper, the subslice path connection problem is modeled as the generalized tsp problem. Ant colony algorithm is used to obtain the optimal solution, and then the starting point and the end point between the sub-slice spray gun paths are connected to form a complete spray gun path and generate the intermediate language program. Compared with the previous heuristic search algorithm such as genetic algorithm, ant colony algorithm has faster convergence speed and better ranking optimization result, that is, the length of connection path is shorter. These advantages are more obvious when the number of spraying paths increases. Finally, the results of spraying path planning and parameter optimization are verified by simulation for typical workpieces. This paper has completed the research and implementation of the job planning method for the complex free-form surface spraying robot. The method can plan the spray gun path for the complex free-form surface, and it is verified by the simulation of the spraying job planning system. Be able to meet project requirements well.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP242

【相似文献】