基于BP神经网络对电弧传感积分差值法的改进
发布时间：2019-07-01 14:16
【摘要】：电流左右区间积分差值法是一种常用的摆动电弧传感偏差提取算法。为了推出电流积分差与位置偏差的关系,需要建立准确的电弧传感数学模型,根据试验得到焊枪高度与送丝速度、焊接电流的线性关系式。针对在焊枪高度很低或很高时线性化处理有较大误差或很难建立准确的电弧传感数学模型的问题,提出了一种基于BP神经网络的方法建立模型并优化积分差值偏差提取算法。进行了多组不同偏差值的焊接试验,分别用2种方法分析处理电流信号,将得到的偏差值与实际偏差对比,推出了线性回归方程并建立、验证了神经网络模型。结果验证了BP神经网络方法的可靠性。
[Abstract]:The current and right interval integral difference method is a common swing arc sensing deviation extraction algorithm. In order to develop the relationship between the current integral difference and the position deviation, an accurate mathematical model of electric arc sensing is needed, and the linear relationship between the height of the welding gun and the wire feeding speed and the welding current is obtained according to the test. In order to solve the problem of large error or difficult to establish an accurate electric arc sensing mathematical model when the height of the welding gun is very low or very high, a method based on the BP neural network is proposed to set up the model and to optimize the integral difference deviation extraction algorithm. In this paper, a series of welding tests with different deviation values were carried out. The current signal was analyzed by two methods, and the obtained deviation value was compared with the actual deviation. The linear regression equation was introduced and the neural network model was established. The results verify the reliability of the BP neural network method.
【作者单位】： 天津工业大学天津市现代机电装备技术重点试验室;解放军汽车修理厂;
【基金】：国家自然科学基金资助项目(U1333128) 天津市科技计划资助项目(14ZCDZGX00802)
【分类号】：TG409
本文编号：2508549
[Abstract]:The current and right interval integral difference method is a common swing arc sensing deviation extraction algorithm. In order to develop the relationship between the current integral difference and the position deviation, an accurate mathematical model of electric arc sensing is needed, and the linear relationship between the height of the welding gun and the wire feeding speed and the welding current is obtained according to the test. In order to solve the problem of large error or difficult to establish an accurate electric arc sensing mathematical model when the height of the welding gun is very low or very high, a method based on the BP neural network is proposed to set up the model and to optimize the integral difference deviation extraction algorithm. In this paper, a series of welding tests with different deviation values were carried out. The current signal was analyzed by two methods, and the obtained deviation value was compared with the actual deviation. The linear regression equation was introduced and the neural network model was established. The results verify the reliability of the BP neural network method.
【作者单位】： 天津工业大学天津市现代机电装备技术重点试验室;解放军汽车修理厂;
【基金】：国家自然科学基金资助项目(U1333128) 天津市科技计划资助项目(14ZCDZGX00802)
【分类号】：TG409
【相似文献】
相关期刊论文 前1条
1 曾智英;张华;叶艳辉;吴恙;;改进的积分差值法在焊枪工作角检测中的应用[J];焊接技术;2013年04期
，本文编号：2508549
本文链接：https://www.wllwen.com/kejilunwen/jiagonggongyi/2508549.html
