检索结果-内蒙古大学图书馆

A data-driven model for predicting metal ductile fracture at various stress states

ENGINEERING FAILURE ANALYSIS 2025年 175卷

作者： Liu, Wen Chen, Geng Liu, Jianhui Zhang, Lele Lanzhou Univ Technol Sch Mech & Elect Engn 36 Pengjiaping St Lanzhou 730050 Gansu Peoples R China Beijing Jiaotong Univ Sch Mech & Elect Control Engn 3 Shangyuan Village Beijing 100044 Peoples R China Beijing Jiaotong Univ Natl Int Sci & Technol Cooperat Base 3 Shangyuan Village Beijing 100044 Peoples R China

Metal ductile fracture often occurs in stamping forming and collision accidents, and accurate prediction of material fracture is essential in the design of stamping parts and passive safety structures. The existing fracture models are based on the general fracture law of materials and can predict fracture in all stress states with fewer parameters. However, the fracture prediction at certain stress states is not accurate enough due to configuration limitations. To this end, a datadriven fracture model was proposed in this paper, which determines the relationship between the fracture strain and the stress triaxiality entirely based on the material fracture behaviour. Thus, high accuracy of fracture prediction under all stress states can be achieved, but the model has more parameters. To calibrate these parameters, tensile experiments were conducted with a series of notched specimens, and the optimization framework of parameters inverse identification was established, including optimization problem establishment, parameter normalization, transformation to an unconstraint problem and solving with the simplex method. The calibration results show that the simulation error of the data-driven model is reduced from 30.64 % to 10.59 % in the 59 iterations, much lower than 18.20 % of the Johnson-Cook model. In the simulation of the modified compact-tension specimen, loads decrease earlier in the Johnson-Cook model, which further highlights the ability of the data-driven model describing the fracture under complex stress states. The accurate fracture prediction of the data-driven fracture model provides a basis for the reasonable design of stamping parts and passive safety structures.

关键词： Data-driven fracture model Stress state Johnson-Cook model Inverse parameter identification framework optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Back analysis of mechanical parameters based on GPSO-BP neural network and its application

引用

SCIENTIFIC REPORTS 2025年第1期15卷 1-20页

作者： Shi, Song Miao, Yichen Di, Cheng Zhao, Quanchao Zheng, Yantao Liu, Changwu Sichuan Univ Coll Water Resource & Hydropower Chengdu 610065 Peoples R China Kunming Univ Sci & Technol Fac Civil Engn & Mech Kunming 650500 Peoples R China China Railway Eryuan Engn Grp Co Ltd Chengdu 610000 Peoples R China

Rock mass mechanical parameters are essential for the design and construction of underground engineering projects, but parameters obtained through traditional methods are often unsuitable for direct use in numerical simulations. The back analysis method, based on displacement monitoring, has emerged as a new approach for determining rock mass parameters. In this study, an experimental scheme was developed using orthogonal and uniform experimental designs to obtain training samples for the neural network. A GA-PSO-BP neural network model (GPSO-BP) was proposed, combining the fast convergence of the particle swarm optimization (PSO) algorithm and the global optimization capability of the genetic algorithm (GA). This model was applied to invert the rock mass parameters E, mu, phi, and c for deep-buried tunnels. The results indicate that the GPSO-BP neural network model outperforms the BP, GA-BP, and PSO-BP neural network models in terms of faster convergence and higher accuracy. It also shows superior performance in handling small datasets and complex problems, achieving better data fitting and the highest score in rank analysis. The DDR curve further confirms the GPSO-BP model's computational efficiency. When the rock mass parameters derived from this model are applied to forward numerical simulations, the average error across four monitoring projects is only 4.34%, outperforming the other three models. Thus, this study provides an effective method for improving the accuracy of rock mass parameter inversion in underground engineering.

关键词： Rock mass parameters BP neural network Back analysis optimization algorithm Displacement prediction

来源：评论

学校读者我要写书评

暂无评论

Synthesize multiple V/H directional beams for high altitude platform station based on deep-learning algorithm

引用

SCIENTIFIC REPORTS 2025年第1期15卷 1-22页

作者： Mahmoud, Korany R. Montaser, Ahmed M. Helwan Univ Fac Engn Elect & Commun Dept Cairo 11795 Egypt Natl Telecommun Regulatory Author NTRA Giza 12577 Egypt Sohag Univ Fac Technol & Educ Elect Engn Dept Sohag 82524 Egypt

This paper investigates the integration of High-Altitude Platform Stations (HAPS) with Deep Learning (DL) models to enhance coverage capabilities. Recognizing the inherent limitations of traditional HAPS coverage, which is typically confined to a circular area, this work proposes a novel approach utilizing a 60-element Concentric Circular Array (CCA) operating at 2.1 GHz. To dynamically generate multiple vertical/horizontal (V/H) directional beams, the system integrates a Deep Neural Network (DNN) with a modified version of the Gravitational Search algorithm and Particle Swarm optimization (MGSA-PSO) algorithm. This hybrid approach optimizes the feeding phases of the CCA elements, enabling the system to effectively cover diverse road paths. Furthermore, the study incorporates realistic scenarios by utilizing the Computer Simulation Technology-Microwave Studio Suite (CST) with the Earth Explorer (EE) user interface tool to model real-world road paths, including those traversing challenging terrains such as rugged deserts with mountain chains and forested areas.

关键词： Deep learning Deep neural network optimization algorithm Antenna array Beamforming technique

来源：评论

学校读者我要写书评

暂无评论

A hybrid model for diabetic retinopathy and diabetic macular edema severity grade classification

引用

INTERNATIONAL JOURNAL OF DIABETES IN DEVELOPING COUNTRIES 2025年 1-15页

作者： Sabeena, A. S. Jeyakumar, M. K. Noorul Islam Ctr Higher Educ Dept Comp Sci & Engn Kanyakumari 629180 Tamil Nadu India Noorul Islam Ctr Higher Educ Dept Comp Applicat Kanyakumari 629180 Tamil Nadu India

BackgroundDiabetic retinopathy (DR) is a common reason for permanent vision loss, particularly among the elderly population worldwide;diabetic macular edema (DME) is characterized by the accumulation of fluid or swelling in the macula, which can occur in any phase of DR progression. New studies have deepened our knowledge of the needs in eye care to improve methods for recognizing, addressing, diagnosing, and treating retinal *** work intends to present a scheme for disease severity classification following the processes: pre-processing, segmentation, and disease severity classification. Firstly, pre-processing is done by employing a median filter, followed by GD-STFA-based SwinUNet for segmentation of lesions into soft and hard exudates, hemorrhages, and ***, DR and DME severity classification is employed by AlexNet-DQN, which is proposed by combining AlexNet and DQN models. Moreover, the proposed AlexNet-DQN is trained using the Exponential Gradient Descent-Sea Turtle Foraging algorithm (ExpGD-STFA), which is developed by including the exponential weighted moving average (EWMA) concept in *** presented mechanism attained better results when compared to traditional *** DR severity classification, the proposed mechanism attained an accuracy of 98.3%, sensitivity of 98.9%, specificity of 97.8%, and precision of 97.6%. Meanwhile, for DME severity classification, the presented mechanism attained an accuracy of 98.1%, sensitivity of 98.8%, specificity of 97.5%, and precision of 97.3%.

关键词： Diabetic retinopathy Diabetic macular edema Segmentation Severity grade classification optimization algorithm Deep learning

来源：评论

学校读者我要写书评

暂无评论

A deep learning based model for diabetic retinopathy grading

引用

SCIENTIFIC REPORTS 2025年第1期15卷 1-20页

作者： Akhtar, Samia Aftab, Shabib Ali, Oualid Ahmad, Munir Khan, Muhammad Adnan Abbas, Sagheer Ghazal, Taher M. Virtual Univ Pakistan Dept Comp Sci Lahore 54000 Pakistan Appl Sci Univ Coll Arts & Sci Comp Sci Dept POB 5055 Manama Bahrain Natl Coll Business Adm & Econ Sch Comp Sci Lahore 54000 Pakistan Korea Univ Dept Biostat Coll Med Seoul 02841 South Korea Gachon Univ Fac Artificial Intelligence & Software Dept Software Seongnam 13557 South Korea Prince Mohammad Bin Fahd Univ Dept Comp Sci Al Khobar 34754 Dhahran Saudi Arabia Al Ahliyya Amman Univ Hourani Ctr Appl Sci Res Dept Networks & Cybersecur Amman Jordan

Diabetic retinopathy stands as a leading cause of blindness among people. Manual examination of DR images is labor-intensive and prone to error. Existing methods to detect this disease often rely on handcrafted features which limit the adaptability and classification accuracy. Thus, the aim of this research is to develop an automated and efficient system for early detection and accurate grading of diabetic retinopathy severity with less time consumption. In our research, we have developed a deep neural network named RSG-Net (Retinopathy Severity Grading) to classify DR into 4 stages (multi-class classification) and 2 stages (binary classification). The dataset utilized in this study is Messidor-1. In preprocessing, we have used Histogram Equalization to improve image contrast and denoising techniques to remove noise and artifacts which enhanced the clarity of the fundus images. We applied data augmentation techniques to our preprocessed images in order to tackle class imbalance issues. Augmentation techniques involve flipping, rotation, zooming and adjustment of color, contrast and brightness. The proposed RSG-Net model contains convolutional layers to perform automatic feature extraction from the input images and batch normalization layers to improve training speed and performance. The model also contains max pooling, drop out and fully connected layers. Our proposed RSG-Net model achieved a testing accuracy of 99.36%, specificity of 99.79% and a sensitivity of 99.41% in classifying diabetic retinopathy into 4 grades and it achieved 99.37% accuracy, 100% sensitivity and 98.62% specificity in classifying DR into 2 grades. The performance of RSG-Net is also compared with other state-of-the-art methodologies where it outperformed these methods.

关键词： Diabetic retinopathy Convolutional neural network Deep learning Augmentation optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Improved Predictive Current Controller for Variable Flux Memory Machine Drives Considering Magnetization State Manipulating Operations

引用

IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS 2025年第2期13卷 1406-1417页

作者： Liu, Xing Yang, Hui Cai, Shun Lin, Heyun Yu, Feng Yang, Yong Southeast Univ Sch Elect Engn Nanjing 210096 Peoples R China Southeast Univ Jiangsu Prov Key Lab Smart Grid Technol & Equipmen Nanjing 210096 Peoples R China UCL Dept Mech Engn London E20 2AE England Nantong Univ Sch Elect & Informat Engn Nantong 226019 Peoples R China Soochow Univ Sch Rail Transportat Suzhou 215131 Peoples R China

Variable flux memory machine (VFMM) is recognized as a competitive counterpart to traditional permanent magnet synchronous machine (PMSM) for electrified transportation, due to the advantages in terms of extended speed range and improved global efficiency. Regarding the control aspect of VFMM drives, the most crucial challenge refers to the design of current controller with emphasis on the exclusive magnetization state (MS) manipulation, which is realized via short d-axis current pulse injection. This article proposes an improved finite-control-set model-free predictive current control (FCS-MFPCC) for VFMM drives. In order to ensure simultaneously the dq-axis current tracking performances during MS manipulating processes, a novel hybrid optimization algorithm is proposed. In particular, a sliding mode-based regulation dedicated to the q-axis current is additionally designed in the proposed algorithm, and by doing so, it is guaranteed that the q-axis current is regulated toward its reference. Finally, experiments on a VFMM test rig are conducted to validate the effectiveness and superiority of the proposed FCS-MFPCC method.

关键词： Current control Robustness optimization Prediction algorithms Torque Prototypes Predictive models Power electronics Mathematical models Heuristic algorithms Finite-control-set model-free predictive current control (FCS-MFPCC) optimization algorithm permanent magnet synchronous machine (PMSM) variable flux memory machine (VFMM)

来源：评论

学校读者我要写书评

暂无评论

OR based multi-objective decision-making in e-commerce improvement

引用

SERVICE ORIENTED COMPUTING AND APPLICATIONS 2025年 1-11页

作者： Gollavilli, Venkata Surya Bhavana Harish Nagarajan, Harikumar Alagarsundaram, Poovendran Gattupalli, Kalyan Sitaraman, Surendar Rama Khalid, Haris M. Under Armour Baltimore MD 21202 USA Global Data Mart Inc GDM South Plainfield NJ USA Humetis Technol Inc Kingston NJ USA Yash Tek Inc Mississauga ON Canada Intel Corp Santa Clara CA USA Univ Dubai Coll Engn & Informat Technol Dubai 14143 U Arab Emirates

The inefficiencies and high costs of traditional business process redesign techniques, which frequently rely on manual optimization and lack systematic approaches, are discussed in this study. Our suggestion to address these issues is a computational fuzzy model that combines multi-objective decision-making methods with operational research. This model analyses corporate goals, resources, and limitations in an organized manner to improve the efficacy of business redesign. Our approach makes finding the best redesign tactics easier while considering various goals and trade-offs. Test results show that the suggested methodology delivers an astounding 98.62% average accuracy in redesign results. Its efficiency is further demonstrated by the fact that the average time needed for the redesign process is reduced to 11.55 s, and related expenditures are kept to a minimum. The results show that using the computational fuzzy model makes decision-making more accessible and generally raises the standard of business operations. This model offers a valuable tool for businesses looking to improve operational effectiveness and save expenses in fast-paced e-commerce by providing an organized approach to company redesign. The outcomes show how sophisticated decision-making frameworks may be integrated into business process management to enable more efficient and flexible organizational procedures.

关键词： Decision-making optimization algorithm e-commerce Business process Multi-objectives

来源：评论

学校读者我要写书评

暂无评论

A model for speed and fuel refueling strategy of methanol dual-fuel liners with emission control areas

引用

TRANSPORT POLICY 2025年 161卷 1-16页

作者： Gao, Tianhang Tian, Jia Liu, Changjian Huang, Chuan Wu, Hongyu Yuan, Ziwen Minist Transport Div Waterway Planning Transport Planning & Res Inst Beijing 100028 Peoples R China

Methanol dual-fuel (DF) liners can simultaneously use traditional fuel oil and methanol as blended fuels, which can address environmental protection requirements while ensuring economy. This paper presents an optimization model for the speed and refueling strategy of methanol DF liners. A decision-making tool is introduced for shipping companies to develop liner operation plans. A mixed-integer 0-1 planning model is constructed with the objective of minimizing the operating cost of a single voyage of a methanol DF liner, and the optimal solutions for the voyage speed in each segment and the fuel oil and methanol refueling strategy at each port are computed considering the dual-fuel mixing of fuel oil and methanol. In this paper, the validity of the model and algorithm are verified with the AEU3 route of COSCO Shipping as an example, and the results show that installing a scrubber on liners and mixing heavy fuel oil (HFO) and methanol is more economical than mixing very low-sulfur fuel oil (VLSFO) and methanol, considering the existing emission standards. The studied liner should increase its speed in emission control areas (ECAs) and refuel at ports with low prices for fuel oil and methanol. Although the above conclusions are not influenced by changes in the price of methanol, the price difference between HFO and VLSFO, or the carbon allowance price, changes in sulfur emission standards will have a significant effect on the speed and refueling strategy of liners and carbon emissions. This paper provides a theoretical reference for operational decision-making for shipping companies operating methanol DF liners and is of practical value for improving the scientific management of methanol DF liners, enhancing the energy efficiency of ships, and reducing the emission of pollutants from ships.

关键词： Methanol dual-fuel liners Speed and refueling strategy optimization algorithm Emission control areas

来源：评论

学校读者我要写书评

暂无评论

Measurement of four main catechins content in green tea based on visible and near-infrared spectroscopy using optimized machine learning algorithm

引用

JOURNAL OF FOOD COMPOSITION AND ANALYSIS 2025年 138卷

作者： Luo, Wei Li, Wenyoujia Liu, Shuling Li, Qicheng Huang, Haihua Zhang, Hailiang East China Jiaotong Univ Coll Elect & Automat Engn Nanchang 330013 Peoples R China Jiangxi Inst Sci & Technol Informat Nanchang 330046 Peoples R China

The catechin content is critical for assessing the nutritional value and commercial worth of green tea, making detection methods valuable for its quality evaluation. This study utilized visible and near-infrared spectroscopy (Vis-NIRS, 400-2500 nm) combined with an optimized machine learning model, integrating the global search whale optimization algorithm (GSWOA) and kernel extreme learning machine (KELM), to predict the content of four main catechins (EC, ECG, EGC, and EGCG) in fresh tea leaves. Multiple preprocessing and effective wavelength selection methods were applied to improve prediction accuracy. The results showed that the distribution of effective wavelengths verified the importance of the visible spectrum. The GSWOA-KELM demonstrated superior performance, optimizing the modeling capabilities of KELM and outperforming other models, including partial least squares regression (PLSR) and random forest (RF), with R2P values exceeding 0.98 and RMSEP values below 3.3 for all catechins. Notably, even without additional data processing, the model maintained good effects, achieving a minimum R2p of 0.9751 and a maximum RMSEP of 3.3654. Therefore, the proposed approach proved to be both accurate and rapid for non-destructive, online catechin analysis and had practical value for tea appreciation.

关键词： Visible and near-infrared spectroscopy Green tea Catechin content Machine learning optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Detection of Quality Deterioration of Packaged Raw Beef Based on Hyperspectral Technology

引用

FOOD SCIENCE & NUTRITION 2025年第3期13卷 e70022页

作者： Wu, Cheng Feng, Yingjie Cui, Jiarui Yao, Zhang Xu, Hailong Wang, Songlei Ningxia Univ Sch Food Sci & Engn Yinchuan Peoples R China Yinchuan Hitech Ind Dev Zone Yinchuan Peoples R China

It is an important measure to ensure food quality and safety that real-time monitoring of the key quality indicators of fresh meat after packaging in the process of storage and transportation. The feasibility of combining hyperspectral imaging (HSI) technology with chemometrics and deep learning to detect the quality deterioration of polyethylene (PE)-packaged raw beef, especially for a key lipid oxidation indicator of malondialdehyde (MDA) content, was explored in this study. The feasibility of filtering to overcome the interference of packaging film on the spectral data was further investigated. Near-infrared HSI (400-1000 nm) was used to collect spectral and spatial data of beef samples during short-term storage. A least squares regression (PLSR) and echo-neural network optimized by vulture optimization algorithms (BES-ESN) models were developed by multivariate data processing methods. The results showed that the performance of models established by PE-packed beef samples was usually inferior to that established by unpacked beef samples. The changes of MDA content in beef were visualized according to the optimal model. In addition, Gaussian filtering was applied to reduce the interference effect caused by the packaging film. The results have demonstrated that HSI combined with Gaussian filter preprocessing and multivariate data processing provided an efficient and reliable tool for detecting the freshness of beef in PE packaging. The best model had a coefficient of determination (R2P) of 0.8309 and a root mean squared error of prediction (RMSEP) of 0.2180, demonstrating the potential of hyperspectral techniques for real-time monitoring of packaged raw meat quality. The findings can provide some references for the meat industry to develop advanced non-invasive quality assurance systems in the meat industry.

关键词： chemometrics hyperspectral imaging malondialdehyde optimization algorithm packaged beef visualization

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：