摘要
根据大气污染物排放浓度变化特点,将无偏GM(1,1)模型与神经网络模型组合,并以矩阵型输入方式替代传统的数列型数据输入方式,得到改进型灰色神经网络模型,称为UGMN模型。接着,采用烟囱入口烟气自动监控系统(CEMS)数据,将模型运用于贵州省某电厂白天及夜间两段时间段内大气污染物排放浓度的模拟与预测。研究结果表明UGMN模型预测精度较好,可以应用于火电厂大气污染物排放浓度预测。
Based on the variation characteristics of air pollutant concentrations,an improved grey neural network model,the UGMN model,was created by combining an unbiased GM（ 1,1） model with a neural network model,using a matrix input mode instead of the traditional sequence data input method. The UGMN model was applied to the simulation and prediction of both daytime and nighttime air pollutant emissions from a power plant in Guizhou Province,China. Simulation data were gathered from a continuous emission monitoring system（ CEMS） in the chimney inlet. The results showed that the UGMN model had a higher prediction accuracy,and the model could effectively be used for the prediction of air pollutant emissions from thermal power plants.
出处
《环境工程学报》
CAS
CSCD
北大核心
2016年第5期2547-2550,共4页
Chinese Journal of Environmental Engineering
关键词
灰色神经网络
无偏GM(1
1)
火电厂大气污染
排放浓度预测
grey neural network
unbiased GM（1
1）
air pollution of thermal power plants
emission concentration prediction
