一种新型的焊接机器人控制器  

A new improved controller for welding robot system

作　　者：张超[1] 王琦[1] 姚永刚[1] 王鑫[2] 雷晶[2] 

机构地区：[1]河南机电高等专科学校,河南新乡453002 [2]陕西科技大学电信学院,陕西西安710021

出　　处：《电焊机》2013年第6期95-98,共4页Electric Welding Machine

基　　金：大学生创新创业训练计划项目(201210708049)

摘　　要：焊接机器人系统的核心是伺服系统,伺服系统电机位置的精确控制是焊接机器人技术的关键,传统的PID控制算法很难对其进行精确的自适应跟踪控制。介绍了一种改进型的神经动态规划(NDP)算法,其中模型模块用小波神经网络来替代,并针对焊接机器人伺服系统的电机位置进行精确的自适应跟踪控制。仿真结果表明,基于改进型NDP算法的控制器在鲁棒性、控制精度和控制效果都优于传统的PID控制器。Servo system is the most important part of welding robot which the accurate control of motor position is the key.It is difficuh of traditional PID control algorithm to adaptively adopt accurate tracking control on welding robot system.In this paper,an improved neural dynamic programming （NDP） ＇algorithm which the model network is wavelet neural network is introduced to solve the prooblem adaptively adopt accurate tracking control on the motor position of servo system.The results show that improved NDP algorithm has the better robustness, control accuracy and control effects than traditional PID control algorithm.

关 键 词：神经动态规划 伺服系统 焊接机器人系统 

分 类 号：TG409[金属学及工艺—焊接]