摘要
采用多元线性回归和BP人工神经网络两种方法分别建立城市污水排放量的预测模型,并进行实例计算验证。通过比较分析,发现BP神经网络的非线性映射能力能够更好地反映城市污水量与多个自变量间的复杂关系,具有较高的模拟精度且应用简便。
Multivariate linear regression and BP artificial neural network were used respectively to set up the forecasting model for urban wastewater discharge, which was verified through case based calculation. By means of comparative analysis, it is observed that BP neural network has the nonlinear mapped ability to reflect well the complex relationship between urban wastewater discharge and several independent variables, and it has the advantages of high analog accuracy and convenience for application.
出处
《中国给水排水》
CAS
CSCD
北大核心
2005年第9期40-42,共3页
China Water & Wastewater
基金
天津市科技发展计划项目(033113811)
关键词
城市污水排放量
预测
多元线性回归
BP人工神经网络
urban wastewater discharge
forecasting
multivariate linear regression
BP artificial neural network