课程文献中心 更多>>

关键词： Design automation   Modeling   Clouds   Large language models   Sequences   Sequential analysis   Codes   Training   Accuracy   Tuning  

摘要： Point-cloud-based CAD reverse engineering aims to recover editable parametric models from scanned geometry. Existing methods remain limited by inaccurate effective parameter prediction, redundant point-cloud inputs, and costly task-specific training. To address these limitations, we propose LiCoCAD, a lightweight model–LLM collaborative framework that reconstructs point clouds as CADQuery code. It uses lightweight modules to generate key constraints and prompts for LLM-based CAD reconstruction. The lightweight CAD-Coder predicts command sequences as structural constraints, while a training-free geometry keypoint extractor converts scanned point clouds into compact coordinate prompts as geometric constraints. These structural and geometric constraints guide commercial LLMs to reason about CAD parameters and generate CADQuery code. This design avoids task-specific LLM fine-tuning and trains only the lightweight CAD-Coder, thereby shifting the main training burden to lightweight task-specific model training. During inference, a CD-feedback refinement stage enforces code validity and refines geometric parameters. Experiments on three datasets show that LiCoCAD achieves higher reconstruction accuracy and robustness than existing methods. Real-world industrial cases further show that the generated CADQuery code remains editable. These results suggest that the lightweight model–LLM collaborative paradigm provides a scalable solution for CAD reverse engineering and is transferable to broader engineering tasks.

关键词： 产教融合   二维CAD   金教材   教学实践   能力培养  

摘要： 制造业转型升级与产教融合实践的持续深化，推动工程类职业院校优化技术技能人才二维CAD应用能力评价体系。在工业软件国产化政策导向下，本研究围绕行业实际需求，聚焦国产CAD软件教学落地场景，分析其对人才培养体系的优化价值，整合优质教材建设成果与教学方式创新，构建系统化实施方案，以期提升学生工程实操水平及岗位适配度，为职业教育产教融合改革推进及高水平技术技能人才培育提供实践参考。

关键词： CAD/CAM restorative material   Luting Material   Resin-matrix Ceramic   Leucite Reinforced Glass Ceramic   Microhardness  

摘要： This study evaluated the time-dependent surface microhardness of different resin-based luting materials beneath CAD/CAM restorative materials of varying types (resin-matrix ceramic-RMC and leucite-reinforced glass ceramic-LRGC) and thicknesses (1 mm and 2 mm) as an indirect indicator of their polymerization behavior. A RMC(Cerasmart 270, GC) and a LRGC (Initial LRF, GC) CAD/CAM ceramic were used. Four luting materials (light-cured/LC, dual-cured/DC, self-adhesive resin cements/SC, and a flowable composite resin/FCR) were evaluated. Specimens (n = 10 per group) were polymerized under no restoration (control 1 mm and 2 mm thick CAD/CAM materials using an LED curing unit (0 mm distance). Top and bottom surface microhardness values were measured at 1 h, 24 h, 7 days, and 30 days. Data were analyzed using non-parametric statistical tests. Restorative material type, thickness, luting material, and time significantly influenced microhardness values (p < 0.05). Increasing restorative thickness resulted in reduced microhardness values, particularly in LC. LC consistently exhibited bottom/top ratios below the 0.80 threshold under all conditions, whereas DC and FCR maintained values above this threshold. RMC materials yielded higher microhardness values compared to LRGC. Time-dependent increases in microhardness were observed, particularly in DC and SC systems. The microhardness of resin-based luting materials beneath CAD/CAM restorations is influenced by the combined effects of restorative material type, thickness, luting material, and time. Increasing restorative thickness reduced microhardness values, particularly in LC materials, whereas time-dependent increases were most evident in DC and SC materials. Not applicable. This study is a laboratory-based investigation and does not involve a clinical trial. Therefore, no clinical trial number has been assigned.

关键词： CAD/CAM blocks   fatigue   flexural strength   hardness   repolishing   surface roughness   toothbrushing simulation  

摘要： Objective This study aimed to evaluate the effectiveness of periodic repolishing on the surface and mechanical properties of resin-based CAD/CAM materials subjected to combined mechanical cycling and simulated toothbrushing, with or without periodic repolishing. Methods Two resin-based CAD/CAM materials were investigated (Tetric CAD [TC] and Shofu HC [SH]). Groups (n = 12) were assigned according to the evaluation period and repolishing (RPL) condition. The aging protocol comprised simulated toothbrushing (ST-200 g load, 250 cycles/min) combined with mechanical cycling (MC-50 N load, 1.67 Hz frequency) performed in aging intervals (AI-2.4 & times;10(5) MC + 10,000 ST totaling two periods (5 & times;AI and 10 & times;AI). Periodic repolishing was performed using a clinical polishing system (OptraGloss, Diamond High-Gloss polisher [HP]). Surface roughness (Ra Vickers microhardness (VH flexural strength (FS and elastic modulus (EM) were quantitatively evaluated, while SEM and AFM analyses were performed qualitatively. Data normality and homogeneity were verified, followed by two-way ANOVA and Tukey's post hoc test (alpha = 0.05). Results Aging significantly increased surface roughness and reduced microhardness in both resin-based CAD/CAM materials (p < 0.05). Periodically repolished groups (T-1RPL and T-2RPL) showed higher VH and lower Ra values (p < 0.05) than their non-repolished counterparts (T-1 and T-2). No significant differences in FS or EM were observed after aging (p = 0.664 and 0.185, respectively). Conclusion Aging significantly affected the surface roughness and microhardness of both materials. Periodic repolishing proved to be an effective strategy for minimizing the effects of aging, thereby reducing surface roughness and preserving microhardness. Clinical Significance Intraoral repolishing at yearly recall intervals may help control surface degradation, maintain smoother surfaces, and preserve microhardness over time, particularly for indirect restoratio

关键词： Optimization   Coronary heart disease   Deep learning   Stenosis   Machine learning   Diagnostic medicine   Cardiovascular diseases   Preprocessing  

摘要： Coronary Artery Disease (CAD) is a leading cause of mortality worldwide and is primarily associated with atherosclerotic plaque formation, resulting in coronary artery stenosis. The accurate prediction of CAD using deep learning models is often constrained by the limitations of conventional optimization techniques, including premature convergence and limited adaptability of the models. To address these challenges, this study proposes a Dynamic Optuna Hybrid Harris Hawks Optimization (Dynamic Optuna H-HHO) framework to enhance the performance of deep learning-based CAD prediction models. The proposed approach integrates dynamic parameter adjustment, adaptive escape energy mechanisms, and Optuna-based hyperparameter tuning. The framework was applied to optimize several deep-learning classifiers, including ResNet-50, VGG-16, InceptionV3, and MobileNet, using a coronary artery stenosis dataset. The performance was evaluated through a comparative analysis with models optimized using the conventional Hybrid Harris Hawks Optimization (H-HHO) algorithm. The experimental results indicate that the proposed Dynamic Optuna H-HHO framework consistently improves the predictive accuracy across all evaluated models. InceptionV3 achieved the highest accuracy of 97.9%, followed by MobileNet with 97.6%, compared with the maximum accuracy of 82.46% obtained using traditional HHO-based optimization. By combining adaptive optimization strategies with automated hyperparameter tuning, the proposed framework provides a robust and scalable solution for improving the accuracy of coronary artery disease prediction.

关键词： Modeling   Remote sensing   Signal detection   Cams   Computer aided manufacturing   Transformers   Convolutional neural networks   Visualization   Dies   Buildings  

摘要： Change detection (CD) methods based on Mamba and state space models have attracted attention for their capacity to capture long-range spatial correlations with linear-time efficiency. However, most existing Mamba-based CD approaches use uniform scanning from generic vision backbones, which treats all spatial regions equally. This overlooks the sparsity and geometric diversity of real changes, resulting in diffused change responses. Driven by these limitations, we propose CAM-CD, a change-aware Mamba framework that approaches CD through sequential modeling. Unlike uniform scanning, CAM-CD aligns state space modeling with sparse, structured changes by guiding state transitions toward change-relevant regions. Specifically, a Change-Aware Adaptive Scanning (CAAS) module utilizes spatial change priors to guide the scanning process, suppressing background noise with a small parameter increase. In addition, a Selective Scanning Fusion Module (S2FM) models bi-temporal feature interaction as a path-dependent sequential processing mechanism, where recursive hidden states accumulate change-relevant information to gather discriminative evidence and enhance long-range contextual understanding. Comprehensive evaluations on four benchmark datasets show that CAM-CD consistently yields superior results compared to contemporary CNN-, Transformer-, and Mamba-based methods. In particular, we achieve IoU gains of 1.09%, 0.68%, 2.28%, and 1.97% over the leading competitors on the LEVIR-CD, WHU-CD, SYSU-CD, and CLCD datasets. Code is available at https://***/xjkgis/CAM-CD.

关键词： Roads   Transportation   Design automation   Road transportation   Modeling   Design methodology   Geometry   Databases   Road safety   Context  

摘要： The characterization of horizontal alignment is essential for road safety analyses and operational performance assessment; however, the availability of detailed geometric data remains limited in many contexts. In this scenario, open geospatial data sources, such as those provided by OpenStreetMap (OSM), have been widely used in transportation studies, although their reliability for applications requiring high geometric precision is still under investigation. This study aims to comparatively evaluate the consistency of OSM-derived horizontal alignment parameters relative to a georeferenced CAD reconstruction. The analysis was conducted on a segment of the BR-116/SP highway, where simple horizontal curves were identified and parameters such as curvature radius and arc length were extracted. The comparison between methods was performed using statistical error metrics (Mean Absolute Error, Root Mean Square Error and Mean Absolute Percentage Error) and graphical analysis, including scatter plots and error distribution. The results indicated different levels of agreement between the evaluated parameters. Radius estimates exhibited stronger association with the reference reconstruction (R2 = 0.749), whereas arc length estimates showed more moderate agreement (R2 = 0.473). These findings suggest that OSM-derived geometry is more effective in reproducing curve radius than arc length in the analyzed highway segment. For radius estimation, a mean error of -12.14 m and a standard deviation of 108.61 m were observed, indicating a tendency toward underestimation and considerable variability among curves. Percentage errors were more pronounced for curve length, suggesting limitations in the geometric delineation of curved segments. It is concluded that, although OSM shows potential for exploratory and large-scale analyses, its use in applications requiring high geometric precision should be approached with caution, and validation with reference data is recommended.

摘要： 近日，西安工程大学、东华大学、太原理工大学等单位共同承担的“废旧涤/棉纺织品高值化利用关键技术及其装备研制”项目通过科技成果鉴定，项目成果达到国际先进水平，其中废旧棉基纸纱、高纤维体积含量废旧纤维复合板材制备技术达到国际领先水平。我国是全球第一纺织大国，每年产生废旧纺织品约2000万吨，但综合利用率不足15%。大量废旧纺织品仍以填埋、焚烧方式处置，造成资源严重浪费与环境污染，对节约资源、减污降碳构成制约，严重影响国家“双碳”战略实施与纺织行业绿色低碳转型。

关键词： angiogenesis   antibacterial   chick embryo   chitosan composites   natural polymer   wound-healing  

摘要： Chitosan extracted from Labeo rohita fish scales via chemical demineralization, deproteinization, and deacetylation yields 25% overall from raw scales. The chitosan has 88%-90% deacetylation, a molecular weight (MW) of 109.5 kDa, and low residual ash (0.1%) and protein (0.29%). Composite films (CGP) with fixed gelatin: polyvinylpyrrolidone (PVP, 10:5) and varying chitosan (1%-7% w/v) are prepared. FTIR confirms cross-linking;SEM shows smooth surfaces for CGP 3% and CGP 5%, but rougher texture for CGP 7%. Water uptake decreases with higher chitosan content (160% for CGP 1% vs. 22% for CGP 7%). PBS buffer uptake is highest for CGP 1%-5% (80%-120% and lowest for CGP 7% (similar to 40%). Lower chitosan content slows degradation. Tensile strength increases with chitosan concentration (27 MPa for CGP 1% to 32.8 MPa for CGP 7% while elongation at break decreases (13% to 4%). Antibacterial activity against Escherichia coli and Staphylococcus aureus is strongest for CGP 3% and CGP 5% (inhibition zones 13-14 mm outperforming pure chitosan. In the hen's egg test on the chorioallantoic membrane (HET-CAM) assay, CGP 3% induces the highest angiogenic response (27.2 +/- 0.3 vs. 12.9 +/- 0.6 vessels in control with no toxicity. CGP 3% and CGP 5% are optimal for wound-healing applications.

关键词： CAD   GS   GWAS   Heat tolerance   Urechis unicinctus  

摘要： Urechis unicinctus is an emerging aquaculture species with high economic value. Restricted by temperature conditions, it is mainly distributed in the intertidal zone of Bohai Sea and parts of the Yellow Sea in China. In addition, high temperatures in summer often cause massive mortality of U. unicinctus, which severely hinders the development of its aquaculture industry. Identifying SNPs and candidate genes related to heat tolerance is crucial for breeding heat-tolerant varieties of U. unicinctus. In this study, genotyping was performed using whole-genome resequencing. After quality filtering, a total of 17,693,070 high quality SNPs were obtained for genome wide association analysis (GWAS). A total of 209 heat tolerance related SNPs and 318 candidate genes were identified. Fine-mapping was used to analyze significant heat tolerant SNPs and identify causal variants. Sanger sequencing was performed on a validation population to confirm the genotyping results. By integrating GWAS and transcriptome data, six genes (MYT1, CAD, SMC6, SPG7, KLHL15, HEATR5B) containing significant heat tolerant SNPs were determined to be associated with heat tolerance. Furthermore, RNAi confirmed the critical role of CAD in heat resistance. The heritability estimate for heat tolerance was 0.376 ± 0.124, indicating a moderate level. Genomic selection (GS) analysis indicated that RR-BLUP was the optimal model for heat tolerance traits. These findings provide a theoretical basis for understanding the molecular mechanisms underlying heat tolerance in U. unicinctus, and offer technical support for marker assisted selection and breeding of heat tolerant varieties, which will accelerate the sustainable development of its aquaculture industry.