基于神经网络反演温度场的燃机余热锅炉汽包有限元分析  

Finite Element Analysis of Waste Heat Boiler Based on Inversion Temperature Field Using Neural Network

作　　者：张毅乐 傅洪军 王瑾 俞洪锋 叶笃毅[1] ZHANG Yi-le;FU Hong-jun;WNG Jin;YU Hong-feng;YE Du-yi(Institute of Process Equipments,Zhejiang University;Wenzhou Gas Turbine Power Generation Co.,Ltd.)

机构地区：[1]浙江大学化工机械研究所 [2]温州燃机发电有限公司

出　　处：《化工机械》2022年第1期39-45,共7页Chemical Engineering & Machinery

基　　金：国家自然科学基金项目(51675475)。

摘　　要：以某9E燃机联合循环发电机组余热锅炉汽包为对象,提出了一种基于神经网络结合机组运行监控数据反演汽包内壁换热系数的方法。并以反演的热边界条件为基础,开展了汽包有限元瞬态温度场模拟,得到汽包特定部位温度的模拟值与实测值基本一致。利用反演温度场进一步开展了汽包瞬态应力场分析,获得了热启动过程汽包关键部件最大等效应力的时间变化历程。Through taking the waste heat boiler of a 9E combustion engine which combined cycle generating unit as the object of study and basing on combing the neural network with the monitoring data,a method to reconstruct heat transfer coefficient of the boiler’s inner wall was proposed.Based on the inverse thermal boundary conditions and adopting the finite element method,simulating the steam drum’s transient temperature field was performed to obtain a good agreement between temperatures simulated and values monitored at the specific location of the drum,including applying the inverse temperature field to analyze boiler drum stress field to acquire the time course of boiler drum key components’maximum equivalent stress during the hot-start process.

关 键 词：汽包 余热锅炉 换热系数 神经网络 温度场 应力场 反演 有限元 

分 类 号：TQ052.6[化学工程]