Agent任务可视编程和单元组合重构

VISUAL PROGRAMMING FOR AGENT TASKS AND CELL RECONSTRUCTION

文中将机器人装配单元的各类设备定义为独立的Ａｇｅｎｔ，采用ＰｅｔｒｉＮｅｔ模型，为各Ａｇｅｎｔ的作业定义了有关操作的基本模型结构，作为基于多Ａｇｅｎｔ的模型库．通过对Ａｇｅｎｔ模型库元素的可视化组合重构，不仅可以实现机器人作业单元几何元素的重组，而且可以方便地实现作业任务的重组．文中提出的基于层次化节点及节点槽结构，实现了对可视基元的分类管理和组织，解决了可视图符与操作任务间的映射问题，可为用户提供基于直接操作的可视化任务程序制作界面．通过采用基于面向目标层次化可视基元的可视交互流程导引，定义了Ａｇｅｎｔ间的作业关系，最终完成对于用户单元组合布置与重构过程的导航。

Each device in the robot assembly cell is defined as a dependent agent in the paper. By means of Petri net models, basic modeling structures for operations of each agent task are defined as the modeling library based on multi\|agent. By visually reconstructing agent model elements, it is possible to realize not only the reconstruction of geometrical elements in the robot cell but also the reconstruction of working tasks. A structure based on layered nodes and node slots is presented and applied to organize and manage visual units according to their classes. The structure is applied to solve mapping problems between visual marks and their operation tasks, and supply a visual task programming design interface based on directly operation for users. With the navigation of visual interactive scheme based on object\|oriented layered visual units, the relations among agents are defined and the guidance of layout and reconstruction processes for user cells is finally completed. Thus, a robot cell visual layout and reconstruction system and a robot cell task visually programming platform are implemented.