Wuhan Technical College of Communications

Archive numbering is a core identification system for the management of academic resources in universities, which directly affects the efficiency of resource retrieval, shelving and utilization. Due to differences in knowledge classification logic, management systems and technological application backgrounds, Chinese and American universities have formed distinctly different archive numbering rules. Taking China's Library Classification for Chinese Libraries (LCCL), archive numbering rules and America's Library of Congress Classification (LCC), university archive numbering systems as the research objects, this paper compares and analyzes their differences from four dimensions: compilation principles, structural systems, coding rules and application characteristics, explores the cultural, institutional and technological drivers behind these differences, and puts forward optimization ideas of integration and reference. It aims to provide a reference for the standardization and internationalization of university archive numbering systems.

Power data auditing is the cornerstone of a reliable and efficient modern power system. Various deep learning models have been successfully applied to fraud detection in power audit data. However, most of these methods rely on manual trial-and-error and expert knowledge to design the neural architectures and hyper-parameters. To address this limitation, this paper proposes an innovative automated deep learning approach for fraud detection model design using genetic algorithm (GA)-based neural architecture search (NAS), termed NAS-FD. In NAS-FD, convolutional neural network (CNN) is employed as the primary detection model, leveraging its strong data learning and feature extraction capabilities. First, an effective encoding scheme is developed to represent the nueral architectures and hyper-parameters of CNN, as these parameters significantly influence the detection performance. Then, considering detection performance as the objective function, well-designed GA-based evolutionary operations are implemented to optimize the neural architectures and hyper-parameters of CNN, obtaining the optimized CNN. The detection performance of the proposed NAS-FD method is validated using an electricity theft dataset from the power auditing domain. Experimental results demonstrate that NAS-FD achieves superior detection performance compared with manually designed deep learning models in terms of four performance indices including accuracy, precision, recall, and F1-score.

Accurate detection of structural damage is essential for ensuring the safety and reliability of bridges. However, traditional vibration-based approaches often struggle to capture rich feature representations and adequately model spatial dependencies among sensors. This study proposes a novel bridge damage detection framework, LGSTA-GNN, which integrates local–global spatiotemporal learning with graph neural networks. The framework first extracts multi-scale temporal–frequency features using a multi-scale feature extraction module. A local graph feature extraction module then models intrinsic spatial relationships through graph convolutions, while a global graph attention module adaptively captures inter-sensor dependencies by emphasizing structurally informative nodes. A benchmark dataset generated from a scaled bridge model under progressive damage states is used to evaluate the proposed method. Extensive experiments demonstrate that LGSTA-GNN outperforms multiple graph neural network variants and conventional deep learning techniques, achieving superior accuracy, precision, recall, and F1-score. The confusion matrix and t-SNE visualization further verify its enhanced discriminative capability and robustness. Ablation studies confirm the contribution of each module, highlighting the effectiveness of global attention in identifying subtle structural deterioration. Overall, LGSTA-GNN provides an effective and interpretable solution for intelligent bridge damage detection, with strong potential for practical structural health monitoring and real-time safety assessment.

This exploration introduces a data-oriented scheme for projecting the punching shear resistance (Vn) linked to failure modes (FMs) in slab-column connections with shear reinforcement. Because reinforced concrete (RC) slab-column connections exhibit a simple construction approach, slabs rest directly on columns without beams. In slab-column connections with shear reinforcement, Vn linked to FMs has only very seldom been stated in research up to this point using machine learning approaches. Two accepted and reliable models were considered for estimation: random forests (RF) and adaptive neuro-fuzzy inference system (ANFIS). Nine input parameters about the punching shear mechanism are found using 327 test results from a computational database. The dataset's learning set (70%) and evaluation set (15%) were used to construct, validate, and test the suggested scheme. Its accuracy is greatly affected by the RF and ANFIS hyperparameters, which need to be selected using metaheuristic enhancement tactics. For this, the Prairie Dog Algorithm (PDA) is used. Based on the logic and assessment of analytical justifications, it can be concluded that both models are exact and dependable, with ANF-PDO being marginally superior compared to RF-PDO.

Solid Oxide Fuel Cells (SOFCs) possess various prominent advantages such as high overall energy conversion efficiency, pollution-free, and no need for expensive metal catalysts, which enable them to have broad application prospects in different fields. Currently, the primary research directions in SOFC studies are focused on material selection and modification through doping to enhance fuel cell performance. This article moved beyond the conventional research methods of material selection and doping, and explored the improvement of interfacial contact between various components in SOFCs through acid treatment. This article studied whether there were also issues where the electrochemical performance of the fuel cell failed to reach the theoretical value due to poor contact between the electrode and electrolyte in SDC ceramic pieces sintered at high temperatures. This article discussed the specific influencing factors and the impact of interfacial engineering on fuel cell performance. Press NCAL powder and SDC powder into a three-layer structural symmetrical fuel cell of NCAL/SDC/NCAL, and studied whether the surface of the acid treatment electrode affected the performance of the fuel cell. Considering that the concentrated acid would react with NCAL, and the concentration of the acid was diluted and dropped on the surface of the NCAL electrode to see if the performance of the fuel cell was affected. The results showed that dilute nitric acid treatment can effectively increase the porosity inside the electrode, making the fuel cell show better electrochemical performance. At 550°C, the maximum output power of NCAL/SDC/NCAL structure reaches 281 mW cm^-2. The results showed that moderate acid treatment can effectively improve the electrochemical performance of the fuel cell, indicating that effective interface treatment engineering can effectively help high temperature annealing fuel cells to achieve theoretical power.

This study explored how advanced artificial intelligence technologies, particularly convolutional neural networks (CNNs), can be applied to visual communication design, especially for automated poster layout generation. The study innovatively combined CNN technology with design practices, proposing an object detection model to identify and locate poster elements. This approach improved classification and localization accuracy in complex scenarios, laying the foundation for more efficient and precise automated design tools. The research covered experiment design, CNN application, layout model construction, and optimization processes. Aesthetic principles like the golden ratio and rule of thirds were applied to enhance visual appeal. By integrating these strategies, the study achieved a transition from design logic to computational logic, generating diverse design solutions. It offers new perspectives for artificial-intelligence-assisted visual communication design.