摘要
机器人焊接作业的前期阶段主要目的是利用特定的双目视觉系统实时采集数据,然后将焊枪引导至待焊工件的起始位置,为后续的焊缝跟踪做准备。为了保证所采用算法的可靠性和实用性,设计了简单的速度控制器;为了符合工业实际情况,提高伺服过程的快速性,利用神经网络对相应参数与所设计的目标函数进行拟合,最后利用遗传算法筛选出最优解,通过仿真证明了所用算法的有效性。
This article was the preparation stage of robot welding operations, the main purpose was to guide the welding gun to the starting position using binocular visual servo system collect the real-time data of welder pieces and prepare for seam tracking. In order to ensure the reliability and practicality of the algorithm, a simple speed controller was design. At the same time, in order to comply with the industrial reality and to improve servo process,the neural network was used to fit the parameters and the designed objective function. At last, the optimal solution was calculated by using genetic algorithms, and the simulation results proved the effectiveness of the algorithm.
出处
《焊接技术》
北大核心
2014年第4期16-20,5,共5页
Welding Technology
关键词
机器视觉
机器人焊接
神经网络
遗传算法
machine vision
robot welding
neural network
genetic algorithm
