基于果蝇优化广义回归神经网络的径流预测

Runoff Prediction Based on Fruit Fly Optimization Algorithm Optimized GRNN

针对水文径流的不确定性特点,提出一种新的径流预测模型。该模型在广义回归神经网络的基础上,采用了果蝇优化算法。通过该模型对四川省万源市后河径流进行了预测,结果显示改进后的模型预测精度明显提高。

According to the uncertainty characteristics of hydrological runoff, a new optimization method of parame- ters is proposed. Based on GRNN, this method adopts FOA algorithm. Then the so - called FOAGRNN was applied in the runoff of Hou River located in Sichuan province and the results show that the improved model of FOAGRNN has a higher precision.