基于皮尔森相关性分析和BP神经网络的北京城市雾霾治理对策

Beijing Urban Haze Control Strategy Based on Pearson Correlation Analysis and BP Neural Network

为了明确下一阶段(2020-2035年)北京城市雾霾治理的政策路径,本文使用皮尔森相关性分析和状态转移算法的改进STA-BP神经网络,分析了北京城市雾霾的主要影响因素并预测了不同治理政策的效果。汽油消耗量、单位GDP能耗、北京二氧化硫排放量、机动车保有量与北京PM2.5具有极强的相关性。基于STA-BP神经网络的模拟结果显示,下一阶段北京城市雾霾治理应以控制机动车尾气排放为重点,采用京津冀跨区域治理模式。为建立京津冀跨区域雾霾污染治理协作机制,具体措施包括建立区域大气污染治理联合机构、制定统一的空气质量管理目标、建立健全区域大气污染联防联控管理机制、加强大气环境承载力约束、突出分区污染治理重点和构建区域污染治理的科技联动机制。

At present,urban haze pollution control of Beijing has achieved preliminary results.In order to clarify the policy path of Beijing's urban haze control in the next stage from 2020 to 2035,Pearson correlation analysis and improved STA-BP neural network are used to analyze the main influencing factors of Beijing's urban smog,propose different governance schemes and predict the effects of different governance policies based on historical data.The result of Pearson correlation analysis shows that gasoline consumption,energy consumption per unit GDP,sulfur dioxide emissions in Beijing and motor vehicle ownership had strong correlation with PM2.5.The simulation results of improved STA-BP neural network show that the next stage of urban haze control in Beijing should focus on the control of vehicle exhaust emissions and cross regional governance model of Beijing Tianjin Hebei.In order to establish cross-regional haze pollution control cooperation mechanism in Beijing-Tianjin-Hebei,the specific measures include establishing regional air pollution control joint organization,formulating unified air quality management objectives,establishing and improving the regional air pollution joint prevention and control management system mechanism,strengthening the constraints of the atmospheric environment carrying capacity,highlighting the focus of regional pollution control and building regional pollution control technology linkagemechanism.