基于神经网络与LabVIEW平台建立焊锡真空炉粗锡含铅量的软测量模型与在线监测研究

Research on soft measurement model and on-line monitoring of lead content in crude tin in solder vacuum furnace based on neural network and LabVIEW

焊锡真空炉粗锡含Pb量的高低直接关系到焊锡真空炉的生产效率,为了改变目前粗锡含Pb量只能通过人工化验才能得到的现状,实验基于反向传播神经网络(Back-Propagation Neural Network,BPNN)与广义回归神经网络(Generalized Regression Neural Network,GRNN)算法原理,构建了BPNN与GRNN软测量模型并对这两种模型的预测效果进行了对比分析,结果表明基于GRNN的粗锡含Pb量软测量模型具有较高的预测精度。同时,采用虚拟仪器(LabVIEW)中的Matlab Script节点技术,成功开发了基于LabVIEW的粗锡含Pb量监测系统,实现了基于BPNN与GRNN软测量模型的粗锡含Pb量实时在线软预测,运行结果表明所开发的监测系统运行稳定可靠。

The content of lead in crude tin relates directly to the production efficiency of the solder vacuum furnace.In order to change the current status that the content of lead in crude tin can be only determined by manual testing,the soft measurement models based on algorithm principles of back-propagation neural network（BPNN）and generalized regression neural network（GRNN）were constructed.The predicted results of two models were compared and analyzed.The results showed that the soft measurement model of lead content in crude tin based on GRNN exhibited higher prediction accuracy.Meanwhile,the monitoring system of lead in crude tin was successfully developed using the Matlab Script node technology in LabVIEW,realizing the real-time on-line soft prediction of lead content in crude tin by soft measurement model based on BPNN and GRNN.The running results indicated that the developed monitoring system was stable and reliable.