Current data-driven inversion methods based on deep learning (DL) use an end-to-end learning to obtain a mapping relationship from seismic data to the velocity model. These methods require truncating the… Click to show full abstract
Current data-driven inversion methods based on deep learning (DL) use an end-to-end learning to obtain a mapping relationship from seismic data to the velocity model. These methods require truncating the feature maps in the output layer of the network to build the velocity model with a specified size. Therefore, they lack the flexibility of handling output of different sizes, as the network needs to be retrained to achieve reliable results when changing the desired size of the velocity model. Here, to solve the problem of data-driven velocity inversion with variable size of output, a novel sampling matrix is proposed to compress the seismic data to the same size as the model to avoid clipping the feature maps. The compressed seismic data are fed into the designed deep convolutional neural network (CNN) model for training. Here, the network has a multiscale encoder–decoder structure and is composed of several residual blocks. Also, the multiobjective loss functions balance strategy is used to optimize the training process. Utilizing the trained network to test compressed synthetic seismic data of various sizes, the effectiveness of the proposed method for the inversion of variable-size elastic wave velocity models is verified.
               
Click one of the above tabs to view related content.