城市钢厂大气颗粒物环境风险预警方法探索

Exploration of Environmental Risk Warning Method for Atmospheric Particulate Matter in Urban Steel Mills

有效的环境预警体系对城市钢厂的可持续发展至关重要。大气颗粒物是钢铁生产中排放的主要污染物之一,具有排放源数量大、点源面源共存、无组织排放量占比高等特点。以南方某钢铁集团的烧结厂区域为例,使用BP人工神经网络模型,研究城市钢厂的大气颗粒物风险预警方法。模型使用背景值、气象参数、生产参数作为输入层,颗粒物浓度为输出层进行训练。结果显示,训练集和测试集的相关系数R分别为0.996 4和0.995 8,大部分样本的颗粒物浓度预测值误差在±5μg/m^3范围内,BP网络模型的预测精度较高。

An effective environmental risk warning system is crucial to the sustainable development of urban steel mills. The atmospheric particulate matter is one of the major pollutants emitted in the production of steel,with the characteristics of a great many emission sources,the coexistence of point sources and surface sources,and a high proportion of non-organized emissions.In this paper,taking the example of a sintering plant of a steel group in southern China,the BP artificial neural network model is used to study the method of risk warning of atmospheric particulate matter in urban steel mills. The model uses background values,meteorological parameters,and production parameters as the input layer,and the particle concentration as the output layer. The results show that the correlation coefficient（ R） of the training set and test set is 0. 9964 and 0. 9958,respectively.The predicted value of most samples have an error of less than 5 μg/m-3 and the prediction accuracy of BP network model is high.