MES Welding Parameter Prediction Model Of Welding Plant Research And Application

Improving enterprise competitiveness is an important subject in enterprise research,and improving product quality and productivity is the key link to improve enterprise market competitiveness.Among the factors of welding pipe production,the welding parameters are important indexes to measure the quality of welding pipe,and the reasonable collocation of welding pipe parameters has a decisive influence on the welding quality.Research tube welding during the welding process,the change rule of welding parameter,in turn,this law guidance for practical production,it can not only improve productivity,and timely correct the welding process of abnormal data,has certain use value.Through analyzing the precision welding data extracted from MES system of an ERW welding pipe factory,it is concluded that there is a functional relationship between welding current,welding arc voltage,welding voltage and welding speed in the welding data.This functional relationship can be obtained mathematically.According to actual needs,in the process of the experiment for welding current forecast,was established by applying the method of statistical model of stepwise regression,the welding current,welding voltage,welding speed,welding arc voltage between forecast model;The improved BP neural network method is used to predict the forecast object,and the prediction accuracy is compared with the stepwise regression method.Results show that the prediction precision of BP neural network to data is greater than the stepwise regression method,considering the size of the welding current is under the influence of the quality of product,it is recommended to use the BP neural network to prediction of welding data,improving product quality and productivity.