本文根据交替投影的基本思想 ,提出了一种基于径向基函数网络的小波变换模极大值重构方法 .该方法利用径向基函数网络 (RBFN)在其基函数宽度很大时具有较好的线性映射这一特性来加快收敛速度 ,提高重构精度 .它不仅能较好地重构信号的边缘 ,而且能准确地重构出信号的峰值 .大量仿真实验表明 ,该方法对各种信号是普遍适用的 ,具有收敛速度快 ,重构精度高的特点 In this paper, based on the basic idea of ​​alternate projection, a wavelet transform modulus maxima reconstruction method based on radial basis function network is proposed. This method uses radial basis function network (RBFN) Good linear mapping to speed up the convergence speed and improve the reconstruction accuracy.It can not only reconstruct the signal edge well, but also accurately reconstruct the signal peak.Many simulation experiments show that this method for a variety of Signal is generally applicable, with the characteristics of fast convergence and high reconstruction accuracy