基于遗传算法的RBF网络及应用

RBF Network Based on GA and It's Applications

本文采用遗传算法优化了ＲＢＰ神经网络，讨论了用一种新颖的方式对ＲＢＦ进行编码，提出了以权重为因子的适应度函数。将优化后的 ＲＢＦ神经网络用于泛函逼近和对新疆雅马渡站的年径流量进行预测，证明期有效性。

In this article, we used GA to optimize RBFNN, discussed a novel way to code RBFNN and advanced the fitness function based on weights. We applied the improved RBFNN to function approach and to. predict the annual pathway flow of Xingiiang Yamadu state. The result proved the efficiency.