基于省域尺度工业固废产量预测模型研究  

Research on the Prediction Model of Industrial Solid Waste Production Based on Provincial Scale

作　　者：胡翩 陈孝杨[1] 牛经纬 华红梅 HU Pian;CHEN Xiao-yang;NIU Jing-wei;HUA Hong-mei(School of Earth and Environment,Anhui University of Science and Technology,Huainan 2320001,Anhui,China)

机构地区：[1]安徽理工大学地球与环境学院,安徽淮南232001

出　　处：《喀什大学学报》2023年第6期47-51,共5页Journal of Kashi University

基　　金：中国工程院战略研究与咨询院地合作重点项目“安徽省固废综合高效循环利用战略研究”(2021-DFZ-15);安徽省重点研究与开发计划项目“两淮农林固废碳基符合新材料资源化高效利用关键技术集成与示范”(S202104a06020064).

摘　　要：为提高BP神经网络算法在固体废物预测中的精度,提出了一种SSA-BP工业固废预测模型,并选取了GDP、常住人口、工业生产总值、能源生产总量、R&D内部支出、财政支出、人均GDP、第二产业贡献率、能源消费总量等作为影响因素,用灰色关联分析方法分别对4种情景下安徽省固废产量进行了预测.结果表明,R&D内部支出、能源生产总量、生产总值、灰色关联度携带有效信息高;SSA-BP模型的MAE、MAPE、R2值分别为375.81、0.04、0.99,均高于其他机器学习回归模型;在不同情境下,安徽省未来一段时间内工业固废产量整体呈现增长趋势,到2035年产生量变化区间为18254.36~30831.54万t.To improve the accuracy of the BP neural network algorithm in solid waste prediction,an SSA-BP industrial solid waste forecasting model was proposed,and GDP,permanent population,gross industrial product,total energy production,R&D internal expenditure,fiscal expenditure,per capita GDP,the contribution rate of the secondary industry and total energy consumption were selected as influencing factors.The output of solid waste production in Anhui province was predicted by grey correlation analysis under 4 scenarios.The results showed that R&D internal expenditure,total energy production,gross domestic product,and grey correlation carried highly effective information.The MAE,MAPE,and R2 values of the SSA-BP model were 375.81,0.04,and 0.99,respectively,which were higher than other machine learning regression models.Under different scenarios,the overall output of industrial solid waste in Anhui Province will show an increasing trend in the future,with a variation range of 182.5436 to 308.3154 million tons by 2035.

关 键 词：工业固体废物 产量预测模型 BP神经网络 多情景预测 灰色关联分析 

分 类 号：X705[环境科学与工程—环境工程]