机械式冲床停上死点位置的控制策略设计与实现  

Design and Realization of Control Strategy of Mechanical Punch Stop Top Dead Center Position

作　　者：秦文强[1] 何通能[1] 陈德富[1] 李华楠[1] 

机构地区：[1]浙江工业大学信息学院电子信息与智能系统研究所,杭州310023

出　　处：《计算机测量与控制》2016年第4期98-100,104,共4页Computer Measurement &Control

摘　　要：机械式冲床进行冲压操作时,上死点位置的精度是最关键的问题之一;为了提高冲床停上死点位置的精度,研究了冲床刹车后由于惯性而转过的角度与刹车时影响因子的关系,拟合出刹车曲线,利用此曲线保证冲床刹车后能准确停在上死点;介绍数据的获取方法及流程,利用BP神经网络对冲床刹车曲线进行拟合,以均方误差作为衡量拟合效果的评价指标,并与最小二乘法拟合的曲线进行比较;分析两种方法对冲床刹车曲线的拟合效果,均方误差分别为0.006 092 3和1.624 7,实验结果表明,使用BP神经网络对冲床刹车曲线拟合的效果较好。The top dead center position accuracy is one of the most critical issues when a mechanical punch stamping operation. In order to improve the accuracy of punch stop top dead center position, the relationship between the angle of inertia and the influence factors is stud- ied when a punch braked. We {it a punch brake curve and use it to ensure that the punch can accurately stop at top dead center. This paper introduces the method and process of data acquisition, and uses BP neural network to fit the punch brake curve. Using mean square error as the evaluation index, and compared with the least square method. The fitting effect of the two methods to the punch brake curve was analyzed. The mean square error is 0. 006 092 3 and 1. 624 7 respectively. The experiment result shows that using BP neural network to fit the punch brake curve is better.

关 键 词：机械式冲床 上死点 刹车曲线 BP神经网络 

分 类 号：TP272[自动化与计算机技术—检测技术与自动化装置]