北京市PM_(2.5)浓度与工业污染关系分析——基于误差修正模型和混频误差修正模型

Research on the Relationship Between PM_(2.5) Concentration and Industrial Pollution in Beijing——Based on error correction model and mixed-data error correction model

使用DP-DNN估算2008-2019年北京市PM_(2.5)浓度数据,纳入工业领域变量建立误差修正模型和混频数据误差修正模型,分析雾霾成因,倒逼经济发展方式转变与产业结构优化.结果表明:(1)原煤产量、火电发电量的增长加重了空气环境污染;(2)混频模型下,原煤产量、火电发电量具有更强污染效应,同时汽车制造过程也是空气污染的主要原因之一.从模型比较来看(1)两个模型均具有明显的误差修正作用,回归系数分别为-0.45和-0.27.(2)混频数据误差修正模型在保留误差修正模型优点的同时,通过纳入高低频数据、预设权重多项式和模拟最优滞后阶数,具有更强大的数据信息处理能力.针对上述结果,本文提议大力发展新能源、宣传节能减排理念,进一步提升生产工艺、规范生产流程,注重运输、贮藏等环节,并强化行业监督与经营规范.

DP-DNN estimation of PM_(2.5) concentration in Beijing from 2008 to 2019 is used to build an error correction model and a mixed-data error correction model for industrial pollution analysis.The results show that:(1)the growth of raw coal output,thermal power generation aggravates air pollution.(2)In the mixed-data model,the raw coal output and thermal power generation have a stronger pollution effect,and automobile production is also one of the main causes of air pollution.(3)The two models also have obvious error correction effects,and the regression coefficients are-0.45 and-0.27,respectively.(4)While retaining the advantages of the error correction model,the mixed frequency data error correction model has a stronger data information processing ability by including high and low frequency data,preset weight polynomials and simulated optimal lag order.In response to the above results,this article proposes to vigorously develop new energy sources,promote the concept of energy conservation and emission reduction,further improve production technology,standardize production processes,focus on transportation,storage and other links,and strengthen industry supervision and operating standards.