ANFIS系统在凤滩水库流量预报中的运用

Application of ANFIS System in the Prediction of the Flow in the Reservoir

凤滩水库流域非常广阔,横跨湘鄂黔渝3省1市,气象形态复杂,进行流量预报极难。利用该水库流域12个气象台站和230个区域气象站1995-2014年期间降水量资料,把库区流域划分为3个区域面雨量,依据自然地理、行政地理、混合加密区划原则建立了3种面雨量模型,运用ANFIS自适应模糊神经系统进行研究。结果表明,在3种模型中,依自然地理和混合加密区划模型的模拟效果理想,水库流量预报精度分别为97%和83%。

The river basin is very broad,and the weather patterns are complex,and the weather patterns are very complex. In this paper,the reservoir basin 12 meteorological stations and 230 regional meteorological station 1995-2014 during precipitation data,see the reservoir basin is divided as three areal precipitation,according to the natural geography,administrative geography,mixed encryption principle of division three surface precipitation model is established,using adaptive fuzzy neural system is studied. The results show that in the three models,the simulation results are satisfactory,and the accuracy of reservoir prediction is 97% and 83% respectively.