摘要
投影寻踪技术是一种分析处理高维数据的统计方法,文中针对洪水预报的随机性、不确定性以及洪水成因的复杂性等特点,通过研究流域(嘉陵江北碚河段)洪水成因及流域下垫面条件分析,建立基于遗传算法的多输入-单输出的投影寻踪回归模型。研究成果表明,该模型预报精度较高。
The projection pursuit technology is a type of statistical method for analyzing and processing high dimensional data. According to the characteristics such as randomness and uncertainty of flood forecasting and complexity of flood causes, the paper analyzes the flood cause and underlying surface conditions of the interest region, Beibei river section of Jialingjiang river, builds the projection pursuit regressive model with the multi input and single output based on the genetic algorithm method. The research shows that the model prediction has the higher accuracy.
出处
《东北水利水电》
2013年第12期43-47,72,共5页
Water Resources & Hydropower of Northeast China
基金
四川大学青年教师科研启动基金(2012SCU11063)
关键词
洪水预报
投影寻踪
遗传算法
flood forecasting
projection pursuit
genetic algorithm
