水资源利用随机预测模型研究

Research on Stochastic Forecast Model of Water Resources Use

水资源利用量受多项因素的影响 ,是一动态、非平稳的随机过程。摈弃传统的典型代表年和频率分析等用水预测方法 ,采用时间序列分析法 ,在用有序聚类法与秩和检验法分析和处理序列跳跃项 (或剔除突变项 )之余 ,建立了一个含趋势项、周期项和随机项的模型 ,以对水资源利用序列进行预测。最后用所建模型对河南省开封市水资源利用进行预测 ,结果表明该模型合理可行 ,且有较高精度。

Water resources use is affected by many factors, and it is of a dynamic and non_stable stochastic process. Time series analysis method is used instead of traditional methods of representative water year and frequency analysis for water use forecast. After dealing with components of caper and breakthrough by using methods of sequential clustering, order and test analysis, a model including moving trend, periodic and stochastic components is established to forecast water resources use sequence. The model is checked up by comparison water resources use between calculated and observed, and the results show that the model is reasonable and feasible, and it has high precision.