江苏清安河水质污染预测模型的改进分析  

Improved analysis on prediction model showingwater pollution of Qing'an river

作　　者：王启平[1] 叶金杰[1] 谢能刚[1] 

机构地区：[1]安徽工业大学机械工程学院,安徽马鞍山243002

出　　处：《安徽工业大学学报（自然科学版）》2004年第1期36-38,共3页Journal of Anhui University of Technology（Natural Science）

摘　　要：应用BP人工神经网络模型算法,对清安河水质生化需氧量BOD在实验室条件和野外条件下的两种耗氧系数K1之间的非线性关系进行了拟合分析。其误差检验精度比基于线性回归方法及线性拟合方法更高、更能体现出实测数据之间的原始非线性映射规律,并可部分去除偶然因素的干扰。是一种在水质环境数学模型研究领域中值得应用的新方法。Based on the BP manual nerve network model and algorithm, the nonline relation of coefficient of oxygen consumption -(K1) of water quality biochemistry oxygen demand (BOD) between lab and field was analysed comparably. And its testing accuracy of error is much higher and more practicable in primitive nonline mapping law of measured data and in eradicating the interference of contingency partly than line-regress and line-compare methods, and finding that the BP algorithm is a practicable and worthy method in the model research field of water circumstance.

关 键 词：江苏 清安河 水质模型 BOD 耗氧系数 BP神经网络 水质污染 水质预测模型 

分 类 号：X131.2[环境科学与工程—环境科学]