基于机器视觉和PSO的机器人示教路径优化研究

Research on Robot Teaching-playback Path Optimization Method Based on Machine Vision and PSO

工业机器人在处理新的生产任务时,需要对其进行示教。目前,传统的机器人示教再现方法存在示教系统复杂、示教时间长以及操作者需要特殊训练等诸多问题。针对上述问题,该文提出了一种基于机器视觉和人手演示的机器人示教方法。首先通过Microsoft Kinect2.0传感器提取人手的示教路径,然后通过粒子群优化算法PSO对人手演示的初始路径进行优化,获得一条优化的机器人运动路径,让机器人完成示教再现。最后,通过实验对基于机器视觉和PSO的机器人示教方法进行验证。实验结果表明,PSO优化后的路径可以有效缩短机器人的运动时间。

Industrial robots need to be taught the steps of completing tasks when dealing with new tasks.At present,the traditional robot teaching reproduction has many disadvantages,such as the complicated teaching system,the high cost,the waste of time,and the special training that operators need.In view of the above situation,a kind of teach-ing-playback based on machine vision and human hand demonstration was proposed in this paper.First the teaching trajectory of the human hand was extracted through the Microsoft Kinect2.0 sensor.Then the initial the human hand demonstration path was optimized by the particle swarm optimization(PSO)algorithm to obtain a better robot motion path,so that the robot can complete the teaching reproduction.Finally,the experiment about robot teaching-playback method based on machine vision and PSO was carried out.The experiment results showed that the optimized path can effectively shorten the robot's motion time.