| 670t/h煤粉炉飞灰含碳量的神经网络预测建模 |
| |
| 引用本文： | 陈彪,丁艳军,吴占松. 670t/h煤粉炉飞灰含碳量的神经网络预测建模[J]. 电站系统工程, 2005, 21(6): 17-19,22 |
| |
| 作者姓名： | 陈彪 丁艳军 吴占松 |
| |
| 作者单位： | 陈彪（清华大学）;丁艳军（清华大学）;吴占松（清华大学） |
| |
| 摘 要： | 对煤粉炉中影响飞灰含碳量的主要因素进行分析,并根据正交实验原理,对某台670t/h的燃煤锅炉飞灰含碳特性进行多工况热态实验,采用基于LM(Levenberg-Marquardt)算法的BP神经网络建立了锅炉飞灰含碳量的神经网络预测模型,实验验证结果表明该算法不仅收敛速度快,而且模型能根据各种操作参数准确地预报锅炉在不同工况下飞灰含碳量。
|
| 关 键 词： | 锅炉 飞灰含碳量 神经网络 LM算法 优化 |
| 文章编号： | 1005-006X（2005）06-0017-04 |
| 收稿时间： | 2005-05-13 |
| 修稿时间： | 2005-05-13 |
Artificial Neural Network Modeling for Forecasting Carbon Content in Fly Ash from 670t/h Utility Boiler |
| |
| CHEN Biao,DING Yan-jun,WU Zhan-song. Artificial Neural Network Modeling for Forecasting Carbon Content in Fly Ash from 670t/h Utility Boiler[J]. Power System Engineering, 2005, 21(6): 17-19,22 |
| |
| Authors: | CHEN Biao DING Yan-jun WU Zhan-song |
| |
| Abstract: | Main factors effected the carbon content in fly ash were analyzed.With the help of orthogonal experiment theory,carbon content in fly ash of a 670t/h utility boiler was experimentally investigated.Taking advantage of BP neural network based on Levenberg-Marquardt(LM)algorithm,the prediction model of unburned carbon content is established and verified.The results illustrates that the LM algorithm has a rapid convergent speed and the model can exactly forecast the carbon content according to different operating parameters. |
| |
| Keywords: | utility boiler carbon content in fly ash neural network LM algorithm optimization |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
