基于ANFIS模型的年径流预报方法

Annual Runoff Forecast Based on ANFIS

分析以往年径流预报方法的特点,阐述自适应神经模糊推理系统(adaptive network-based fuzzyinference system,ANFIS),提出年径流预报的ANFIS模型,并将其应用到西北地区某水文站年径流预报中.以MATLAB为工具,依据该地区历年水文资料,对年径流量进行预报.实例结果表明,与改进的ANN模型(最速下降—共轭梯度法、进化单纯形法)相比,本方法计算速度快、泛化能力强、预报精度高,说明ANFIS在年径流预报方面具有良好的适用性.

The past forecast methods for annual runoff are analyzed and ANFIS is explained.This paper applies ANFIS forecast model which is used in the annual runoff forecast of a hydrologic station in northwest zone of China.Based on annual hydrologic data of this zone,the annual runoff is forecasted by MATLAB.The results show that this method accelerates the speed of calculation and improves the capability of accommodativeness and the precision of forecast by comparing this method with developed ANN model(two method...