【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。更多还原

【Abstract】 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.更多还原