The learning algorithm and determination of network parameters of dynamic fuzzy neural network (DFNN) implements Takagi¨CSugeno¨CKang (TSK) fuzzy systems based on extended radial basis function neural networks (RBFNN) are introduced, combing wavelet transform with DFNN, a landslide deformation mornitoring data denoising by wavelet transform to divide noise from the original deformation data to obtain the tendency of deformation, and then predicted by DFNN. The prediction precision before and after denoising show that it is more effectively and more precisely to predict the deformable body deformation after denoising for deformation mornitoring data.