Convolutional Neural Network Formulation to Compare 4-D Seismic and Reservoir Simulation Models

Rollmann, Klaus; Soriano-Vargas, Aurea; Almeida, Forlan La Rosa; Davólio, Alessandra; Schiozer, Denis José; Rocha, Armando Freitas da

doi:10.1109/tsmc.2021.3051649

Cited by 9 publications

(2 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Experimental monitoring of seismic performance of semicircular anti-slide piles is an important means to ensure the safety and stability of bridge structures, and can provide scientific basis and technical support for bridge design and construction [16]. The time-history response analysis of vertically prestressed semicircular anti-slide piles under earthquake action mainly refers to describing the effect by calculating the strain rate and acceleration curves after applying a load in a certain direction [17]. The change of vertical load moment under earthquake action is a function of the bearing capacity and displacement of the pile foundation within a certain period of time.…”

Section: Experimental Monitoring Of Seismic Performancementioning

confidence: 99%

Seismic Performance of Vertically Prestressed Semicircular Anti-Slide Piles Based on Deep Machine-learning Algorithms

Peng,

Wang,

Wang

et al. 2024

Proceedings of the First International Conference on Science, Engineering and Technology Practices for Sustainable Development,

View full text Add to dashboard Cite

Deep machine learning algorithm is a method involving technology, data analysis, statistics, monitoring and so on. Its main goal is to extract, transfer, process and manage information from the data, so as to realize the acquisition and application of knowledge. Deep machine learning algorithms are those machine learning algorithms based on multi-layer neural network structure, with strong learning ability and expression ability. The semicircular vertical prestressed anchor cable anti-slide pile is a new technology for the treatment of large-scale landslides on the bank slopes of Jianghe Reservoir and mountain highways. In this structure, the prestressed steel strand is vertically placed in the anti-slide pile, which can reduce the failure of the prestressed anchor cable in water and the loss of the free length. Its semicircular section fully utilizes the eccentric stress characteristics. It not only enhances the compressive performance of the compression zone of the anti-slide pile, but also improves the weakness of the tensile zone of the anti-slide pile. At the same time, compared to traditional circular piles, the cross-sectional area of semicircular piles is reduced, significantly reducing the amount of materials used, and reducing disasters caused by landslide instability and damage caused by large-scale excavation of slopes. The engineering economy and safety are significant. Through theoretical analysis, this paper constructs key technologies such as structural calculation theory and design methods, and uses model tests to verify their correctness and practicality, thereby revealing the mechanical impact mechanism of semi circular vertical prestressed structures. The experimental data show that the seismic performance of horizontal and vertical prestressed anti-slide piles is different, and the seismic performance of vertical prestressed piles has advantages in the early stage of earthquake, with a change rate of 38.5% in axial force and 20.2% in shear strength. The application of deep machine learning algorithms in the seismic performance of anti-skid piles has potential advantages, which can provide more accurate and efficient prediction and optimization methods to support the seismic design and performance evaluation of sliding piles.

show abstract

Section: Experimental Monitoring Of Seismic Performancementioning

confidence: 99%

Seismic Performance of Vertically Prestressed Semicircular Anti-Slide Piles Based on Deep Machine-learning Algorithms

Peng,

Wang,

Wang

et al. 2024

Proceedings of the First International Conference on Science, Engineering and Technology Practices for Sustainable Development,

View full text Add to dashboard Cite

show abstract

“…For this reason, alternative methods for adding seismic data information to the history-matching procedure have been explored. Recently, Rollmann et al [35] presented a method using a convolutional neural network trained to fit observed seismic history. However, results were only shown in a synthetic case, and the overhead cost of gathering the necessary amount of training data (large amounts of models that need to be classified) as well the time spent in appropriate architecture development (which can be very case-specific) and computational costs associated with the training of the network are still big disadvantages.…”

Section: Introductionmentioning

confidence: 99%

Four-Dimensional History Matching Using ES-MDA and Flow-Based Distance-to-Front Measurement

Barrela,

Berthet,

Trani

et al. 2023

Energies

View full text Add to dashboard Cite

The use of 4D seismic data in history matching has been a topic of great interest in the hydrocarbon industry as it can provide important information regarding changes in subsurfaces caused by fluid substitution and other factors where well data is not available. However, the high dimensionality and uncertainty associated with seismic data make its integration into the history-matching process a challenging task. Methods for adequate data reduction have been proposed in the past, but most address 4D information mismatch from a purely mathematical or image distance-based standpoint. In this study, we propose a quantitative and flow-based approach for integrating 4D seismic data into the history-matching process. By introducing a novel distance parametrization technique for measuring front mismatch information using streamlines, we address the problem from a flow-based standpoint; at the same time, we maintain the amount of necessary front data at a reduced and manageable amount. The proposed method is tested, and its results are compared on a synthetic case against another traditional method based on the Hausdorff distance. The effectiveness of the method is also demonstrated on a semi-synthetic model based on a real-case scenario, where the standard Hausdorff methodology could not be applied due to high data dimensionality.

show abstract

A Graph Convolutional Network Approach to Qualitative Classification of Hydrocarbon Zones Using Petrophysical Properties in Well Logs

Venkateshwaran,

Ramkumar,

Siddiqui

et al. 2024

Nat Resour Res

View full text Add to dashboard Cite

Convolutional Neural Network Formulation to Compare 4-D Seismic and Reservoir Simulation Models

Cited by 9 publications

References 32 publications

Seismic Performance of Vertically Prestressed Semicircular Anti-Slide Piles Based on Deep Machine-learning Algorithms

Seismic Performance of Vertically Prestressed Semicircular Anti-Slide Piles Based on Deep Machine-learning Algorithms

Four-Dimensional History Matching Using ES-MDA and Flow-Based Distance-to-Front Measurement

A Graph Convolutional Network Approach to Qualitative Classification of Hydrocarbon Zones Using Petrophysical Properties in Well Logs

Contact Info

Product

Resources

About