摘要
为更好地控制水质预测精度,在灰色模型的基础上引入3次样条插值函数,在结合灰色模型中的参数信息覆盖和用非线性最小二乘法估计未知参数的基础上,提出了改进的曲线拟合模型。采用该模型拟合共产主义渠监测断面历年来的高锰酸盐指数。结果表明,在控制预测精度的情况下,该模型的拟合结果合理。改进的拟合曲线模型具有一定普遍性,可用于河流和湖泊等水环境因子预测。
The paper is aimed at proposing an evaluation method by introducing the kinds of difference three times＇ function into the grey model by means of directly applying the parameter information to cover with further combination of the non-linear least square method so as to improve the estimation of the water quality. In order to reduce the fitting error and improve the accuracy of the post-discrcte, this paper has adopted a derivative of the inserted value function based on the dynamic modeling of the grey systematic theory that is extensively used in program and project evaluation instead of divided difference. As is known, the traditional water quality forecast methods are in most cases merely a means that acts as a feedback in accordance with the forecasting result with no help to control the water quahty forecast result. However, this paper has proposed a curve fit model by introducing kinds of inserting value function three times and, combining non - linear least square method for estimating the unknown parameters. Monitoring a section of a ditch named the Canal Communism, which contains high manganese sour salt elements over the years, we have taken the ditch as a case study and arrived at the conclusion that the curve fitting model can be improved to meet the needs of the sample. The result also indicates that our improved method is fit for the case rationally. The merit of this method is that it can do proper forecast and control under the known fitting error sum of squares premise with the selection of parameters being simpler and much more accu rate.
出处
《安全与环境学报》
CAS
CSCD
北大核心
2009年第4期71-74,共4页
Journal of Safety and Environment
关键词
环境水利
水质模型
3次样条插值
灰色模型
信息覆盖
environmental water resource
water quality model
kinds of difference three times
grey model
informa tion covering
