期刊文献+

基于改进BP网络的煤粉锅炉飞灰含碳量预测 被引量:2

Forecast of Flying Ash Carbon Content of Pulverized Coal-fired Boiler Based on Modified BP Network
下载PDF
导出
摘要 介绍了改进BP神经网络的原理和建立。采用改进BP神经网络,以煤的全水分、空气干燥基水分、挥发分、灰分、低位发热量、煤粉细度、炉膛空气系数、排烟温度8种影响因素作为输入层的输入,以飞灰含碳量作为输出层的输出,对某煤粉供热锅炉的飞灰含碳量进行了预测。预测值与实测值的最大绝对误差为0.046 8×10-2,最大相对误差不超过3%,该预测方法可行。 The principle and establishment of modified BP neural network are introduced, laking eight kinds of influencing factors including total moisture of coal, air dried basis moisture, air dried basis volatile matter, air dried basis ash, net calorific value, pulverized coal fineness, air coefficient in furnace and flue gas temperature as conjunction points of input layer, and carbon content of flying ash as conjunction points of output layer, the carbon content of flying ash of a pulverized coal-fired boiler is forecasted using modified BP neural network. The maximum absolute error between forecasted value and measured value is 0. 046 8×10^-2 with maximum relative error not exceeding 3%. This forecast method is feasible.
出处 《煤气与热力》 2007年第4期59-61,共3页 Gas & Heat
关键词 改进BP网络 燃煤锅炉 飞灰含碳量 预测 modified BP network coal-fired boiler carbon content of flying ash forecast
  • 相关文献

参考文献3

二级参考文献7

共引文献8

同被引文献15

引证文献2

二级引证文献5

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部