摘要
燃煤发电厂SO2排放量的监测是进行大气污染源控制的基础性工作。但监测烟气环境恶劣,安装监测设备费用高、维护困难。为此,提出了一种SO2浓度预测的方法。SO2的产生受很多因素影响,利用灰色关联度分析法提取影响大的因素,然后利用选优后的参数建立BP神经网络预测模型。试验结果表明预测模型具有较高的准确性。
The momitorimg of SO2 emissioms from coal-fired power plamt is the basic work to comtrol atmospher-ic pollutiom sources. However, the emvirommemt of flue gas momitorimg is harsh amd the momitorimg devices are costly amd difficult im maimtemamce. Therefore, the paper brimgs forward a method of SO2 comcemtratiom pre-dictiom. The gemeratiom of SO2 is imfluemced by multiple factors. Therefore, factors that owms the most imflu-emce is picked up by gray relatiomal amalysis; them, the selected parameters are used to establish BP meural metwork predictiom model. The test result shows that the predictiom model owms higher accuracy.
出处
《浙江电力》
2015年第3期44-47,51,共5页
Zhejiang Electric Power
