According to the effecting factor and the characteristic of urban daily water and the error of line model,we set up the nonparametrie multi regressive model of daily water demand.

• 根据城市用水量的影响因素及特点,针对传统的线性回归模型误差较大的缺点,基于核估计与局部线性估计理论,建立了城市日用水量的非参数回归预测模型。