改进RBF神经网络在降水量预测中的应用

Application of Improved RBF Neural Network in Precipitation Forecast

摘要: 采用遗传算法对RBF网络的隐层节点中心值和宽度进行优化,用梯度下降法求出权值.将其应用于降水量预测中,取得较为满意的结果,证明该算法的有效性.

Abstract: By using genetic algorithm,the hidden nodes center value and width of RBF networks were optimized,and weights value was calculated by gradient descent method.Satisfactory results were obtained in application of rainfall forecast,which proved the effectiveness of this algorithm.