Correlation methods of base-level cycle based on wavelet neural network

Abstract

A method of wavelet neural network (WNN) for correlation of base-level cycle is given. Wavelet neural network is introduced first. Then, a new method of Vectored of well logs is proposed. Through the training which causes the wavelet neural network memories the cycle pattern characteristic of the curve, and using the trained wavelet neural network to identify the memory cycle pattern of the well vectored group to be processed, thus realizes the contrast between the well cycles. Real application indicates that it is highly efficient and reliable in base-level cycle correlation.