机器人作业的离散事件建模与控制方法

针对传统机器人控制中规划器开环的不足, 首先设计了底层的非时间参考的连续路径规划与控制方法, 使规划器的规划值基于系统的状态之上, 避免了基于时间规划的系统由于意外事件造成的损害和任务的重新规划; 在此基础上, 又建立了整个作业任务的离散事件模型, 设计了离散事件路径规划方法及控制综合方法, 实现了事件反馈的闭环, 提高了系统处理不确定事件的能力和作业的自动化程度. 最后两台PUMA560机械手搬运被识别工件的作业验证了本文算法的有效性.

Discrete event modeling and control approach for robotic tasks

In order to conquering the disadvantage of the open loop planning in the traditional robot controller, this paper designs the non-time-based continuous path planning and control method first. It enables the controller's planning value based on the system's states to avoid the damage due to the unexpected events and the replanning of the task in the time-based system. And then we set up the discrete event model for the whole task and provide the discrete event path planning and control synthesis methods. Those methods closed the event feedback loop and improved the system's capability to couple with the unexpected events and the automation degree. The task for two PUMA 560 manipulators to carry the recognized objects verified the effectiveness of this arithmetic finally.