基于主成分分析和贝叶斯正则化方法的神经网络年最大洪峰流量预测模型探讨

Annual Peak Discharge Prediction Model of Neural Network Based on Principal Component Analysis and Bayesian Regulation

针时水文预测建模中输入因子过多而导致神经网络结构规模过大,泛化能力差的问题,利用主成分分析和贝叶斯正则化方法对神经网络进行改进,优化网络结构,从而提高泛化能力。以洮儿河流域镇西站年最大洪峰流量预测为例,研究结果表明,改进的神经网络预测方法与传统的神经网络方法相比,泛化能力有显著提高,而且网络的收敛也比较稳定,实际预测中效果良好。

Aiming at the complex framework of hydrology prediction model of neural network, which leads to decrease the prediction precision, the paper gave a model using principal component analysis and Bayesian regulation. Taking the annual peak discharge at Zhenxi Station as an example, it showed that the method could effectively reduce the size of the model and the generalization capability of the model was better than the traditional neural network.