基于BP神经网络的贴片机运动精度控制  

Movement Precision Control of SMT Placement Based on BP Neural Network

作　　者：刘明晓[1] 龙绪明[1] 

机构地区：[1]西南交通大学电气工程学院,成都610031

出　　处：《计算机测量与控制》2014年第12期3973-3975,3979,共4页Computer Measurement &Control

摘　　要：将BP神经网络PID控制方法应用于贴片机运动精度控制器设计;针对传统PID控制器参数难以整定等问题,提出了BP神经网络和PID控制器相结合的方法,该方法既有常规PID控制器结构简单的特点,又有BP神经网络自适应、自学习以及逼近任意函数的能力;首先根据伺服电动机的工作原理建立了电枢控制伺服电动机模型传递函数,在此基础上建立了贴片机单关节轴位置控制器模型传递函数;其次描述了BP神经网络和PID控制器相结合的控制模型,并对具体的控制算法进行了定义;最后利用MATLAB仿真工具对贴片机单关节位置控制进行了仿真;仿真结果表明,结合BP神经网络的PID控制系统提高了系统的稳定性、快速性和动态性能并获得很好的控制效果。A control method for designing the precision control of SMT Placement based on BP neural network and PID control method.To deal with the traditional PID controller parameters tuning difficult issues,propose a new type of PID control method based on BP neural network,it has the merits of both neural BP network and PID controller.Firstly,the transfer function of armature control servo motor model is established according to the working principle of the servo motor.On this basis,the transfer function of SMT signal joint axis position controller model is established.Second,describes the combination control model of the BP neural network and the PID strategy,and the specific control algorithm is defined.Finally,the MATLAB simulation tools is used to simulate the SMT single joint axis position con trol.The simulation showed that the BP neural network PID control system enhance the stability and rapidly of the system,improve the dynamic performance of the system and get a better results.

关 键 词：贴片机 快速高精度 运动控制 BP神经网络 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]