加热炉炉温神经网络预测模型的开发

Development of Neural Network Prediction Model for Temperature of Heating Furnace

针对加热炉系统非线性、大滞后、大惯性,炉温难以有效预测的问题,以山东钢铁莱芜分公司宽厚板加热炉为研究对象,通过神经网络训练获得充分逼近仿真对象的系统参数,最后使用该方法对莱钢宽厚板加热炉炉温进行预测,结果说明该方法预测准确,具有较强的实践意义,为炉温控制提供了可靠依据,提高了生产效率,降低了能耗。

The temperature inside heating furnace is hard to predict due to nonlinear, high hysteretic and big inertia of the system. Aimed to the wide-heavy plate heating furnace of Laiwu Steel, system parameters fully close to emulated object were obtained through neural network training. Finally this approach was used to predict temperature inside the wide-heavy plate heating furnace of Laiwu Steel, the results of which showed that the method predicts accurately, bears practical significance, provides reliable basis for furnace temperature control, improves production efficiency and reduces energy consumption.