基于长短时记忆神经网络的锌液温度预测模型

Temperature Estimation of Liquid Zinc Based on Long Short Time Memory Neural Network Prediction Model

锌液温度控制是热浸镀锌工艺的难点之一。通过对某铁塔制造企业热浸镀锌工艺分析和镀锌车间现场数据整理,选取影响锌液温度的多个因素,建立了基于长短时记忆神经网络的锌液温度预测模型。实验结果表明,该模型能有效预测未来时刻的锌液温度,为操作人员控制锌液温度提供有意义的参考。

The temperature control of liquid zinc is one of the difficulties in hot dip galvanizing process.In this paper,by analyzing the hot-dip galvanizing process of an iron tower manufacturing enterprise and sorting out the field data of the galvanizing workshop,multiple factors affecting the temperature of liquid zinc are selected,and a prediction model of liquid zinc temperature based on long-term and short-term memory neural network is established.The experimental results show that the model can effectively predict the temperature of liquid zinc in the future,and provide meaningful reference for operators to control the temperature of liquid zinc.