超超临界锅炉高温受热面热偏差预测模型研究

A Thermal Deviation Prediction Model for the High-temperature Heating Surface of Ultra Supercritical Boilers

通过BP神经网络建立锅炉运行方式与热偏差的映射模型,并对热偏差模型的输入变量进行灵敏度分析,找到锅炉热偏差较大工况时控制减少高温受热面热偏差的主要因素,建议以降低管壁温度峰值,减缓氧化皮生成速度,减少超温爆管事故,提高锅炉运行安全性。

A mapping model of boiler operating mode and thermal deviation was established based on BP neural network, while a sensitivity analysis was made on input variables of the model to find the main factors that affect the thermal deviation of high-temperature heating surfaces under the condition of high boiler thermal deviation, so as to provide suggestions to lower the peak temperature of tube walls, slow down the formation rate of oxide scales, reduce the risk of explosive accidents, and therefore improve the safety of boiler operations.