Data Processing of Reflection Seismic Data by Use of Neural Network

The present paper proposes an algorithm for data processing of reflection seismic data by use of neural network. The algorithm of neural network was applied to the reading of the first break, the recognition of waveform in seismic trace and the automatic picking of the result of the constant velocity scan among the various data processing techniques. A layered network with training by the error back propagation algorithm was used. The general procedure of processing of reflection seismic data by use of neural network is as follows. 1) Constitute the most suitable network for the target processing. 2) Set the weight values to all units in the layers and the teacher's signal. 3) Calculate the output signals from the output layer by activating the network. 4) Estimate the learning signal from the actual output signal and the teacher's signal. 5) Calculate the change of weight values so as to minimize the differences between the actual signal and the teacher's signal. 6) Repeat step 3) to 5) till the errors fall into the designated limitation or the designated leaning count is reached. As a result of model studies, it was clarified that the proposed algorithm performed the first break reading, the waveform detection and the automatic picking of velocity with good accuracy.