基于炉区分段的BP神经网络钢坯温度预测

Billet Temperature Prediction Based on Reheating Furnace Zone BP Neural Network

利用神经网络可以逼近任意非线性系统的特点,建立加热炉内钢坯温度预报模型。由于钢坯的边界热流系数和热传导系数、比热、重度等物理参数都是随温度变化的函数,所以,按加热炉的分区采用多个BP神经网络进行分段预测,以适应钢坯的物理参数的变化。仿真结果表明:所建模型简单,精度高,能满足工程实际需要。

Based on the fact that neural network can approximate arbitrary nonlinear system, the article establishes the billet temperature prediction model. Due to heat flux coefficient, heat conduction coefficient, specific heat and density of billet is constantly changed with temperature, so, we adopt multiple BP neural network for temperature predition in order to adapt changes of the physical parameters of billet, according to reheating furnace segment zone. Simulation results show that the model is simple, high precision, corresponded with engineering in practice.