KEYWORDS: 3D modeling, Data modeling, Visual process modeling, Process modeling, Gallium nitride, Convolution, Mathematical modeling, Geology, Visualization, Computer simulations
Automatic history matching is a process of using an optimization algorithm to adjust the parameters of the reservoir model. The reservoir model can reproduce the historical performance of the reservoir and realize the prediction for future production. Accurate prediction of oil well performance guarantees to establish a reliable reservoir model, which is traditionally realized by ESMDA and ensemble Kalman filter. We design and implement history matching using a 3D-pix2pix generative adversarial network(3D-pix2pix GAN) structure for the first time, which can correct the parameters of the complex heterogeneous reservoir based on dynamic response. The adversarial generative network includes generator and discriminator. The generator attempts to use the fast feedforward operation of historical production data (input) to reconstruct the calibrated model, while the discriminator attempts to distinguish the pseudo output and the prior (real data) so that 3D-pix2pixGAN finally learns an infinitely close to the real reservoir model. The most significant contribution of this work is to train a 3D-pix2pixGAN model to correct reservoir model parameters. Compared with traditional work ow, 3D-pix2pixGAN has several advantages. First, the reservoir parameters estimated from history matching help to improve 3D reservoir characterization. Second, the reservoir obtained by history matching can accurately predict the future production of water and oil. Third, 3D-pix2pixGAN is used as a proxy model instead of using a numerical simulator in the training process to reduce the amount of computation and achieve end-to-end offline processing.
A great number of associative memory models have been proposed to realize information storage and retrieval inspired by human brain in the last few years. However, there is still much room for improvement for those models. In this paper, we extend a binary pattern associative memory model to accomplish real-world image recognition. The learning process is based on the fundamental Hebb rules and the retrieval is implemented by a normalized dot product operation. Our proposed model can not only fulfill rapid memory storage and retrieval for visual information but also have the ability on incremental learning without destroying the previous learned information. Experimental results demonstrate that our model outperforms the existing Self-Organizing Incremental Neural Network (SOINN) and Back Propagation Neuron Network (BPNN) on recognition accuracy and time efficiency.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.