基于BP网络的水污染物排放预测模型研究  

Research on Predictive Model of Water Pollution Emissions Based on BP Neural Network

作　　者：王之仓[1] 俞惠芳[1] 

机构地区：[1]青海师范大学计算机学院,青海西宁810008

出　　处：《青海师范大学学报（自然科学版）》2012年第4期13-16,共4页Journal of Qinghai Normal University(Natural Science Edition)

基　　金：教育部创新团队项目(No.IRT1068);青海省科技厅应用基础研究(No.2011-Z-750);青海省科技厅软项目(No.2009-Z-719);青海师范大学科研创新项目

摘　　要：由于青海省西宁市区和大通县排污、城镇污水处理设施建设相对滞后和工业企业废水达标排放率较低,湟水河小峡桥断面等6个断面连续5年各水期水质均劣于Ⅴ类.黄河中上游流域控制单元青海段总共有5个控制单元,其中2个位于湟水河.基于此,本文建立了通用的研究企业的能源消耗数据与水污染物排放之间的非线性关系的模型,发现了湟水河流域大型企业能耗和水污染排放物之间的非线性关系,预测了湟水河流域大型企业水污染物排放量,对保护黄河上游水环境具有非常重要的现实意义.The water quality in every water-stage is inferior to V class at 6 sections containing Xiaoxia Bridge section of the Yellow River for 5 consecutive years, the reasons are that pollution discharge of Xining city and Datong county of Qinghai province pollution discharge is serious, sewage treatment facilities of city and town are lagging and Industrial wastewater has a low emissions standards. There are 5 control units at the Qinghai Stage of the upper reaches in the Yellow River, and 2 control units locate in Huangshui River. Therefore, based on the theory of artificial neural network, we build a general non-linear relation model of energy consumption and water emissions about enterprises and find the relationship of Huangshui River basin, then we predict the water pollution emissions of large-scale enterprises. The fact shows that it has important practical significance to protect the environment of Yellow River＇upstream.

关 键 词：BP网络 能源消耗 水污染 非线性 COD 氨氮 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]