基于GASS-BPEE交叉训练算法的河道冰情预报组合模型

Combination model of river ice regime forecast based on GASS-BPEE cross-training algorithm

在文章的河道冰情预报的组合模型中,首先在成因分析的基础上,采用逐步回归算法对预报因子进行有效的筛选,然后采用GASS-BPEE交叉训练算法对冰情要素进行预报。利用该组合模型对凌汛灾害多发的松花江依兰、佳木斯江段开河日期进行了预报,结果表明所建立的河道冰情预报组合模型结构简单、预报精度较高,具有实用价值。

For the combination model of river ice regime forecast,the forecast factors are effective selected firstly by using the stepwise regression algorithm based on the basis of the cause analysis,then the factors of the ice regime are forecasted by using the GASS-BPEE cross-training algorithm.The combination model is used to forecast the break-up time of the Yilan and Jiamusi section of the Songhua River,with more ice flood disaster,the results show that the model has the advantages of the simple structure and high forecasting accuracy,and has practical value.