基于BP神经网络-系统动力学耦合模型的江苏省水资源承载力预测与调控研究  

Study on forecast and regulation of water resources carrying capacity in Jiangsu Province based on GA-BP-SD coupling model

作　　者：薛晴 杨侃[1] XUE Qing;YANG Kan(College of Hydrology and Water Resources,Hohai University,Nanjing 210098,Jiangsu,China)

机构地区：[1]河海大学水文水资源学院,江苏南京210098

出　　处：《水利水电技术（中英文）》2022年第11期86-99,共14页Water Resources and Hydropower Engineering

基　　金：山西省水利科学技术研究与推广资助项目(2017DSW02);国家重点基础研究发展计划(973计划)资助项目(2012CB417006);国家科技支撑计划资助项目(2009BAC56B03)。

摘　　要：为量化调控指标对水资源承载力未来发展态势的影响程度,以优化配置水资源并充分发挥其综合效益,以江苏省为研究对象,构建系统动力学(SD)预测模型,对现状延续下2019-2030年江苏省水资源承载力进行动态模拟。在此基础上耦合遗传算法(GA)改进的Back Propagation人工神经网络(ANN)模型,对水资源承载力进行评分。同时进行障碍度诊断,分析在未来社会经济发展中影响江苏省水资源承载力的障碍因子,依此筛选调控指标。运用情景分析法,针对调控指标设置五种调控方案,分析污废水排放总量、氨氮排放量、化学需氧排放量(COD)、生产用水量和居民生活用水量5个反映用水水量和排放水水质的量质要素的变化趋势以及水资源承载力评分走势,结果表明:江苏省水资源承载力不断恶化,较2019年、2030年水资源承载力评分下降9.61%,氨氮排放量提高6.82%,总用水量提高23.94%,难以满足未来经济社会发展的需要。方案1、方案2及方案4从降低用水总量角度出发,方案3从控污角度出发,都有效改善了水资源承载力,但均无法逆转水资源承载力在2019-2030年的下降态势;方案5统筹考虑节流、调整产业结构和水污染处理调控措施,较2019年、2030年水资源承载力评分提高3.15%,氨氮排放量下降44.13%,总用水量下降9.90%,有效缓解水资源供需压力、改善水环境质量,实现了江苏省水资源承载力在未来年份中的稳步提升,为优化水资源调控提供建设性依据。In order to quantify the degree of influence of control indicators on the future development trend of water resources carrying capacity to optimize the allocation of water resources and give full play to its comprehensive benefits,taking Jiangsu Province as the research object,this paper builds a system dynamics(SD)prediction model to dynamically simulate the water resource carrying capacity of Jiangsu Province from 2019 to 2030 under the continuation of the current situation.On this basis,coupled with the Back Propagation artificial neural network(ANN)model improved by genetic algorithm(GA),the water resources carrying capacity is scored.At the same time,the obstacle degree model is introduced to diagnose the obstacle factors that will affect the water resources carrying capacity of Jiangsu Province in the future social and economic development,and to screen the control indicators accordingly.Scenario analysis is used to set five kinds of regulation for regulation program indicators.The change trend of the total amount of waste water discharge,ammonia nitrogen discharge,chemical oxygen discharge,production water consumption and household water consumption,five quantitative and qualitative factors that reflect the amount of water used and the quality of discharged water and the scoring trend of water resources carrying capacity is analysed.The results show that the water resource carrying capacity of Jiangsu Province is deteriorating continuously.Compared with 2019 and 2030,the water resource carrying capacity score decreases by 9.61%,ammonia nitrogen emission increases by 6.82%,and total water consumption increases by 23.94%.It is difficult to meet the needs of future economic and social development.Scheme 1,Scheme 2,and Scheme 4 are from the perspective of reducing the total water consumption,and Scheme 3 is from the perspective of pollution control.They have effectively improved the water resources carrying capacity,but none of them can reverse the decline of the water resources carrying capacity in 2018-2030.

关 键 词：水资源承载力 系统动力学 BP神经网络 障碍度模型 调控指标 情景分析 

分 类 号：TV213.4[水利工程—水文学及水资源]