Search

This paper proposes a parallel computation model for the self-organizing map (SOM) neural network applied to Euclidean traveling salesman problems (TSP). This model is intended for implementation on the graphics processing unit (GPU) platform. The Euclidean plane is partitioned into an appropriate number of cellular units, called cells, and each cell is responsible of a certain part of the data and network. Compared to existing GPU implementations of optimization metaheuristics, which are often based on data duplication or mixed sequential/parallel solving, the advantage of the proposed model is that it is decentralized and based on data decomposition. Designed for handling large-scale problems in a massively parallel way, the required computing resources grow linearly along the problem size. Experiments are conducted on 52 publicly available Euclidean TSP instances with up to 85,900 cities for the largest TSPLIB instance and 71,009 cities for the largest National TSP instance. Experimental results show that our GPU implementations of the proposed model run significantly faster than the currently best-performing neural network approaches, to obtain results of similar quality. [ABSTRACT FROM AUTHOR]

In recent years, igneous reservoirs have been the focus of natural gas exploration and development, which can help solve the energy demand, and in addition, when oil and gas reservoirs are depleted, volcanic reservoirs provide good storage space, transport channels and storage capacity, and this is very beneficial for CO2 storage. For reservoir prediction and CO2 storage monitoring, fluid discrimination techniques are very important. For igneous rock, the lithology and mineral composition, porosity, and pore structural characteristics are known to all affect the distribution patterns and saturation of fluid. The presence of high resistivity water layers and low resistivity gas layers will affect the accuracy of fluid discrimination using resistivity data. In this study, based on core experiments, the G-B-P (Gassmann-Brie-Patchy)equation was adopted to describe the relationships between the acoustic velocity and fluid saturation values. Then, finite element acoustic logging numerical simulations of wells were carried out to obtain the array logging full waveforms of the different lithology, porosity, and saturation. Finally, the ICEEMDAN(Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise) mode decomposition method and SPWVD were used to analyze the time-frequency characteristics of the simulated waveforms, and a fluid discrimination chart of gas layer was established. The results of the synthetic and field data experiments using the proposed method demonstrated that it was a good approach for fluid discrimination processes in igneous gas reservoirs. • ICEEMDAN-SPWVD time-frequency analysis. • Full waveforms simulation based on Gassmann-Brie-Patchy equation. • A Fluid Discrimination Method based on array acoustic logging. [ABSTRACT FROM AUTHOR]