SEISMIC INVERSION USING ERROR-BACK-PROPAGATION NEURAL NETWORK

In connection with inversion problem restrained by well log information,a large network system is built in this paper to approach the direct modeling and inversing process of the inverse problem on condition that a nonlinear map exists between the inverse goal and the seismic data.The energy function of the system is given in the light of inverse problem, and the error back propagation learning algorithm is used to realize adaptively extrapolatting well log information,so the neural network is Combined with inverse problem better here. The result of the extrapolation of well log velocity using actual information shows: this method has good applied prospects.