Search

The Chang 7 member of the Triassic Yanchang Formation in the Ordos Basin is a significant continent shale oil reservoir in China. Therefore, conducting an in-depth investigation into the pore structure and fluid mobility characteristics of the Chang 7 shale oil reservoir holds immense importance for advancing shale oil exploration. This study conducts a detailed analysis of the pore structures and their impact on fluid mobility of the Chang 7 shale oil reservoir using multiple methodologies, i.e., a cast thin section, scanning electron microscopy (SEM), X-ray diffraction (XRD), high-pressure mercury injection (HPMI), low-temperature nitrogen adsorption (LTNA), and nuclear magnetic resonance (NMR). The results show that the sandstone in the Yanwumao area of the Chang 7 shale oil reservoir consists mainly of lithic arkose and feldspathic litharenite, which can be classified into three lithofacies (massive fine-grained sandstone (Sfm), silt-fine sandstone with graded bedding (Sfgb), and silt-fine sandstone with parallel bedding (Sfp)). Moreover, three pore structures (Type I, II, and III), and four pore spaces (nanopores, micropores, mesopores, and macropores) can be characterized. Pore structure Type I, characterized by large pores, exhibits bimodal pore diameter curves, resulting in the highest levels of movable fluid saturation (MFS) and movable fluid porosity (MFP). Pore structure Type II demonstrates unimodal pore structures, indicating robust connectivity, and higher MFS and MFP. Pore structure Type III primarily consists of dissolved and intercrystalline pores with smaller pore radii, a weaker pore configuration relationship, and the least fluid mobility. Furthermore, a correlation analysis suggests that the pore structure significantly impacts the fluid flowability in the reservoir. Favorable petrophysical properties and large pores enhance fluid flowability. Micropores and mesopores with high fractal dimensions have a greater impact on reservoir fluid mobility compared to macropores and nanopores. Mesopores mainly control MFS and MFP, while micropores govern the shift from bound fluid to movable fluid states. Among the lithofacies types, the Sfm lithofacies exhibit the highest fluid mobility due to their significant proportion of macropores and mesopores, whereas the Sfgb lithofacies have lower values because they contain an abundance of micropores. The Sfp lithofacies also dominate macropores and mesopores, resulting in medium fluid mobility levels. This study combines lithofacies types, micro-reservoir pore structure characteristics, and mobile fluid occurrence characteristics to better understand the dominant reservoir distribution characteristics of the Chang 7 shale oil reservoirs in the Ordos Basin and provide theoretical information for further optimization of production strategies. [ABSTRACT FROM AUTHOR]

Federated Learning (FL) enables the multiple participating devices to collaboratively contribute to a global neural network model while keeping the training data locally. Unlike the centralized training setting, the non-IID, imbalanced (statistical heterogeneity) and distribution shifted training data of FL is distributed in the federated network, which will increase the divergences between the local models and the global model, further degrading performance. In this paper, we propose a flexible clustered federated learning (CFL) framework named FlexCFL, in which we 1) group the training of clients based on the similarities between the clients’ optimization directions for lower training divergence; 2) implement an efficient newcomer device cold start mechanism for framework scalability and practicality; 3) flexibly migrate clients to meet the challenge of client-level data distribution shift. FlexCFL can achieve improvements by dividing joint optimization into groups of sub-optimization and can strike a balance between accuracy and communication efficiency in the distribution shift environment. The convergence and complexity are analyzed to demonstrate the efficiency of FlexCFL. We also evaluate FlexCFL on several open datasets and made comparisons with related CFL frameworks. The results show that FlexCFL can significantly improve absolute test accuracy by $+10.6\%$ + 10. 6 % on FEMNIST compared with FedAvg, $+3.5\%$ + 3. 5 % on FashionMNIST compared with FedProx, $+8.4\%$ + 8. 4 % on MNIST compared with FeSEM, $+4.7\%$ + 4. 7 % on Sentiment140 compare with IFCA. The experiment results show that FlexCFL is also communication efficient in the distribution shift environment. [ABSTRACT FROM AUTHOR]

With the rapid development of the Internet of Things and the improvement of computing power, edge‐computing becomes an emergency computing paradigm that communicates between the terminal and cloud. One of the most representative works of edge‐computing is to achieve human‐activity recognition at the edge side, as it has lower latency and it could reduce transmission costs, compared with processing at the cloud side. However, existing approaches have many drawbacks: (1) they can merely recognize separated actions as it is incompetence for continuous activity recognition; and (2) they are not robust to action transformation and environmental noise due to the value‐feature‐based matching strategy. In this paper, we propose HCAR, a structure‐feature–based human continuous activity recognition system, which is insensitive to action transformation and environmental noise. Firstly, we leverage word2vec to embed the CSI sequences to CSI value space. Secondly, we select representative features from the embedded vectors and use HMM‐LDA to cluster them into different action categories. Lastly, for each new coming sequence, we calculate the Hellinger distance and bi‐modality coefficient to different categories and then identify the corresponding action(s). We implement HCAR by Intel 5300 NIC to evaluate the activity recognition precision in different cases. The experiments show that HCAR can recognize actions corresponding to the unsegmented CSI sequence with high accuracy, ie, >90%. [ABSTRACT FROM AUTHOR]

The southwestern slope of Lufeng Depression, which is located in the central part of Zhu I Depression is a significant geomorphic unit due to its petroliferous discoveries and tectonic evolutions since early Palaeogene. Based on the comprehensive analysis of cores, wireline logs, seismic volumes, and other related geological and geophysical information, the sequence architecture and depositional evolution was analysed, and their responses to regional tectonic processes and base‐level changes were revealed. The results indicated that two second‐order sequences and six third‐order sequences defined by different types of sequence boundaries are recognized. The second‐order and third‐order sequences were bounded by regional erosional unconformities on the depression margin or slope and unconformities in the central depression with the assist of stacking patterns of wireline logs and depositional characteristics of cores. The second‐order sequence generally consists of two regional transgressive–regressive cycles which can be further subdivided into several local sub‐cycles equivalent to corresponding third‐order sequences which can be identified from wireline logs and cores. The palaeogeomorphology resulted from compound and complex tectonic events in the early Palaeogene plays an important and main role in controlling sequence architecture. In contrast, the base‐level changes largely controls the sequence architecture in the late Palaeogene. The depositional systems within the sequence stratigraphic framework are mainly composed of braided delta deposits and shore deposits, which respectively includes braided channel, subaqueous distributary channel, lacustrine shale, beach bar, maximum flooding surface, shoreface, and foreshore. The base‐level changes, which apparently originated from tectonic uplift and subsidence and sediment supply, controlled the regional transgressions and regressions in sedimentary facies distributions. The development of subsequence (what we call it sand units or sections) from 2,980 to 2,880 in Enping Formation in the Early Palaeogene is obviously influenced by a decrease in tectonic subsidence (rifting stage) and an increase in accommodation space (base‐level rising), while the case is opposite in subsequence 2,850 of Enping Formation. However, the subsequence in Zhuhai Formation in the late Palaeogene from 2,760 to 2,600 was formed in stable tectonic settings (transition stage), and thereby, it was mainly dominated by base‐level changes which were caused by sediment supply and accommodation space changes. [ABSTRACT FROM AUTHOR]