针对传统的BP神经网络存在的不足及局限性,文章提出了利用小波分析和神经网络相结合的方法应用于GPS可降水量预测中。小波神经网络是将小波基函数来替代传统神经网络中的激活函数,它将小波分析和神经网络有机融合在了一起,同时具备小波分析和神经网络的良好特性。通过相同样本数据训练和学习以及对预测结果的对比分析,表明小波神经网络在可降水量预测中比BP神经网络具有更好的容错能力和逼近能力,且其收敛速度快,预测精度高。 In view of the shortcomings and limitations of the traditional BP neural network, this paper proposes the application of the combination of wavelet analysis and neural network in predicting GPS precipitable water. Wavelet neural network is a wavelet basis function to replace the traditional neural network activation function, which combines wavelet analysis and neural network organically together, with the advantages of wavelet analysis and neural network. Through the training and learning of the same sample data and the comparative analysis of the prediction results, it shows that the wavelet neural network has better fault tolerance and approximation ability than the BP neural network in predicting precipitable water, and its convergence speed is fast and the prediction accuracy is high.