检索结果-内蒙古大学图书馆

A method for suppressing spectrum aliasing of off-axis digital holograms based on Kronecker interpolation and backpropagation

引用

OPTICS AND LASERS IN ENGINEERING 2024年 175卷

作者： Guan, Ying Cui, Ze Xu, Zihan Zhou, Wenjing Shanghai Univ Dept Precis Mech Engn Shanghai 200072 Peoples R China

The research proposes a new method to solve the spectrum aliasing and improve the resolution of off-axis digital holograms based on Kronecker interpolation and backpropagation algorithms. The method suppresses the direct current (DC) image and increase the low-frequency information with Kronecker interpolation of the hologram. The size of the Fresnel reconstructed image can be controlled with the reconstruction distance, and this feature is used to select the appropriate reconstruction distance for the Fresnel amplitude reconstruction of interpolated holograms to separate the real and imaginary images. To obtain the complete first-order spectral information for the object, the amplitude of the desired image is obtained with spatial domain filtering and then inverted. Finally, the amplitude and phase of the object are reconstructed according to the angular spectrum reconstruction algorithm. The results show that the method can solve the spectrum aliasing problem of the holograms and thus improve the resolution of off-axis holograms. In addition, the background noise of the reconstructed phase is reduced because the influence of the zero-order spectrum can be completely avoided.

关键词： Kronecker interpolation Off -axis digital holography backpropagation algorithm Spectrum aliasing

来源：评论

学校读者我要写书评

暂无评论

Design of backpropagation Neural Network for Aging Estimation of Electric Battery

引用

SENSORS AND MATERIALS 2023年第4期35卷 1385-1395页

作者： Shin, Kyoo Jae Busan Univ Foreign Studies Dept Artificial Intelligence Convergence 65 Geumsaem ro 485 Beon gil Busan 46234 South Korea

The state of charge (SOC) of an electric vehicle is very important for predicting the remaining battery level and safely protecting the battery from over-discharge and overcharge conditions. In this regard, a neural network (NN) algorithm using backpropagation (BP) has been proposed to accurately estimate the SOC of a battery. Lithium polymer batteries have a nonlinear relationship between their estimated SOC and the current, voltage, and temperature. In this study, a lithium polymer battery with a capacity of 3.7 V/16 Ah was applied. A charge/discharge experiment was performed under constant current and temperature conditions at a discharge rate of 0.5 C. The experimental data were used to train a backpropagation neural network (BPNN) that was used to predict the SOC under charging conditions and the depth of dispatch (DOD) performance under discharge conditions. As a result of the experiment, the error of the proposed BPNN model was found to be 0.22% of the mean absolute error in the discharge DOD and 0.19% of the mean absolute error in the charging SOC at 10, 50, 100, and 150 cycles. Therefore, the high performance of the SOC learning model of the designed BP algorithm was confirmed.

关键词： electric vehicle battery state of charge machine learning methods neural network backpropagation algorithm

来源：评论

学校读者我要写书评

暂无评论

Selection of Network Parameters in Direct ANN Modeling of Roughness Obtained in FFF Processes

引用

POLYMERS 2025年第1期17卷 120-120页

作者： Buj-Corral, Irene Sivatte-Adroer, Maurici Rodero-de-Lamo, Lourdes Marco-Almagro, Lluis Univ Politecn Cataluna Barcelona Sch Ind Engn ETSEIB Dept Mech Engn Ave Diagonal 647 Barcelona 08028 Spain Univ Politecn Cataluna Polytech Sch Engn Vilanova & Geltru EPSEVG Dept Mech Engn Ave Victor Balaguer 1 Vilanova I La Geltru 08880 Spain Univ Politecn Catalunya UPC Barcelona Sch Ind Engn ETSEIB Dept Stat & Operat Res Av Diagonal 647 Barcelona 08028 Spain

Artificial neural network (ANN) models have been used in the past to model surface roughness in manufacturing processes. Specifically, different parameters influence surface roughness in fused filament fabrication (FFF) processes. In addition, the characteristics of the networks have a direct impact on the performance of the models. In this work, a study about the use of ANN to model surface roughness in FFF processes is presented. The main objective of the paper is discovering how key ANN parameters (specifically, the number of neurons, the training algorithm, and the percentage of training and validation datasets) affect the accuracy of surface roughness predictions. To address this question, 125 3D printing experiments were conducted changing orientation angle, layer height and printing temperature, and measuring average roughness Ra as response. A multilayer perceptron neural network model with backpropagation algorithm was used. The study evaluates the effect of three ANN parameters: (1) number of neurons in the hidden layer (4, 5, 6 or 7), (2) training algorithm (Levenberg-Marquardt, Resilient backpropagation or Scaled Conjugate Gradient), and (3) data splitting ratios (70%-15%-15% vs. 55%-15%-30%). Mean Absolute Error (MAE) was used as the performance metric. The Resilient backpropagation algorithm, 7 neurons, and using 55% of training data yielded the best predictive performance, minimizing the MAE. Additionally, the impact of the dataset size on prediction accuracy was analysed. It was observed that the performance of the ANN gets worse as the number of datasets is reduced, emphasizing the importance of having sufficient data. This study will help to select appropriate values for the printing parameters in FFF processes, as well as to define the characteristics of the ANN to be used to model surface roughness.

关键词： FFF FDM Polylactic acid artificial neural networks surface roughness training algorithm number of neurons datasets distribution multilayer perceptron backpropagation algorithm

来源：评论

学校读者我要写书评

暂无评论

The Cascaded Forward algorithm for neural network training

引用

PATTERN RECOGNITION 2025年 161卷

作者： Zhao, Gongpei Wang, Tao Jin, Yi Lang, Congyan Li, Yidong Ling, Haibin Beijing Jiaotong Univ Key Lab Big Data & Artificial Intelligence Transpo Minist Educ Beijing Peoples R China Beijing Jiaotong Univ Sch Comp Sci & Technol Beijing 100044 Peoples R China SUNY Stony Brook Dept Comp Sci Stony Brook NY 11794 USA

backpropagation (BP) algorithm has played a significant role in the development of deep learning. However, there exist some limitations associated with this algorithm, such as getting stuck in local minima and experiencing vanishing/exploding gradients, which have led to questions about its biological plausibility. To address these limitations, alternative algorithms to backpropagation have been preliminarily explored, with the Forward-Forward (FF) algorithm being one of the most well-known. In this paper we propose a new learning framework for neural networks, namely Ca scaded Fo rward ( CaFo ) algorithm, which does not rely on BP optimization as that in FF. Unlike FF, CaFo directly outputs label distributions at each cascaded block and waives the requirement of generating additional negative samples. Consequently, CaFo leads to a more efficient process at both training and testing stages. Moreover, in our CaFo framework each block can be trained in parallel, allowing easy deployment to parallel acceleration systems. The proposed method is evaluated on four public image classification benchmarks, and the experimental results illustrate significant improvement in prediction accuracy in comparison with recently proposed baselines. The code is available at: https://***/Graph-ZKY/CaFo.

关键词： Deep neural network backpropagation algorithm Forward-Forward algorithm Image classification

来源：评论

学校读者我要写书评

暂无评论

A machine learning approach for clustered data

引用

COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION 2025年第2期54卷 406-416页

作者： Wang, Jing Univ Utah Dept Family & Prevent Med Salt Lake City UT 84112 USA

Artificial neural networks (NNs) are a machine learning algorithm that have been used as a convenient alternative of conventional statistical models, such as regression in prediction and classification because of their capability of modeling complex relationships between dependent and independent variables without a priori assumptions about the model form and variable distributions. However, traditional NNs cannot incorporate dependencies of data with a clustering or nesting structure involved in longitudinal studies and cluster sampling. This research is intended to fill this literature gap by integrating the random-effects structure into NNs to account for within-cluster correlations. The proposed NN method incorporating random effects (NNRE) is trained by minimizing the cost function using the backpropagation algorithm combined with the quasi-Newton and gradient descent algorithms. Model overfitting is controlled by using the L2 regularization method. The trained NNRE model is evaluated for prediction accuracy by using the leaving-one-out cross-validation for both simulated and real data. Prediction accuracy is compared between NNRE and two existing models, the conventional generalized linear mixed model (GLIMMIX) and the generalized neural network mixed model (GNMM), using simulations and real data. Results show that the proposed NNRE results in higher accuracy than both the GLIMMIX and GNMM.

关键词： Artificial neural networks backpropagation algorithm Clustered data Generalized linear mixed model Leaving-one-out-cross-validation Overfitting regularization

来源：评论

学校读者我要写书评

暂无评论

Optical modeling, solver, and design of macroscopic single-enantiomer carbon nanotube film and reconfigurable chiral photonic device

引用

CARBON 2025年 235卷

作者： Fan, Jichao Hillam, Benjamin Guo, Cheng Fujinami, Hiroyuki Shiba, Koki Xie, Haoyu Chen, Ruiyang Yanagi, Kazuhiro Gao, Weilu Univ Utah Dept Elect & Comp Engn 50 S Cent Campus Dr Salt Lake City UT 84112 USA Stanford Univ Dept Appl Phys 348 Via Pueblo Stanford CA 94305 USA Tokyo Metropolitan Univ Dept Phys 1-1 Minami Osawa Hachioji Tokyo 1920397 Japan

The interaction of circularly polarized light with chiral matter and functional devices enables novel phenomena and applications. Recently, macroscopic solid-state single-enantiomer carbon nanotube (CNT) films have become feasible and are emerging as a chiral photonic material platform thanks to their quantum-confinement- induced optical properties and facile scalable assembly. However, optical modeling, solver, and device design tools for such materials are non-existent. Here, we prepare macroscopic single-enantiomer (6,5) and (11,-5) randomly oriented CNT films and create an optical material model based on measured experimental optical spectra. We also implement a highly-parallel graphic-processing-unit accelerated transfer matrix solver for general bi-anisotropic materials and layered structures. Further, we demonstrate reconfigurable chiral photonic devices in a heterostructure with phase change materials through machine learning-enabled efficient gradient- based inverse design and optimization. Our developed full stack of a chiral photonic material and device hardware platform and a corresponding high-performance differential-programming-enabled solver opens the door for future chiral photonic devices and applications based on single-enantiomer CNT films.

关键词： Single-enantiomer carbon nanotube Transfer matrix GPU acceleration backpropagation algorithm Phase change material Reconfigurable chiral photonic devices

来源：评论

学校读者我要写书评

暂无评论

Neighbor discovery protocol anomaly-based detection system using neural network algorithm

引用

INTERNATIONAL JOURNAL OF INFORMATION SECURITY 2024年第3期23卷 1787-1803页

作者： Saad, Redhwan M. A. Anbar, Mohammed Manickam, Selvakumar Shaheen, Samir I. Hasbullah, Iznan H. Ibb Univ Fac Engn Dept Elect Engn Ibb 70270 Yemen Univ Sains Malaysia USM Natl Adv Ctr IPv6 Gelugor Penang Malaysia Cairo Univ Fac Engn Dept Comp Engn Giza 12613 Egypt

The exponential increase in Internet-facing devices in the last decade has resulted in IP address exhaustion due to the limitations of the existing IPv4 address space. Therefore, the Internet Engineering Task Force engineered a new version of the Internet protocol known as Internet Protocol Version 6 (IPv6) to resolve the issue. However, IPv6 is highly dependent on the neighbor discovery protocol (NDP), which, unfortunately, has well-known vulnerabilities in its underlying messaging protocol, the Internet Control Message Protocol version 6. So, the NDP flaws leave the IPv6 network open to many security threats and attacks, including man-in-the-middle, spoofing, and denial-of-service attacks, which are the most annoying attack at the network layer. Unfortunately, one of the critical issues plaguing the existing anomaly-based detection system is the effectiveness of detecting NDP-based DDoS attacks, which requires urgent attention. This paper suggests a system to find network traffic patterns that are not normal that are caused by NDP-based attacks. It does this by teaching neural networks how to recognize network attack patterns using the backpropagation algorithm. The proposed system is a big step forward from where the field is now because it uses a complex neural network algorithm to create an NDP anomaly-based detection system. Using a real dataset to test the proposed system's performance shows that it can find NDP anomalies with a 99.95% success rate, a 99.92% precision rate, a 99.98% recall rate, an F1-Score of 99.98%, and a 0.040% false positive rate. Also, the proposed approach shows better results compared to other existing approaches.

关键词： NDP Anomaly detection Neural networks backpropagation algorithm Feature selection IPv6 network

来源：评论

学校读者我要写书评

暂无评论

Exit wave reconstruction of a focal series of images with structural changes in high-resolution transmission electron microscopy

引用

JOURNAL OF MICROSCOPY 2024年第1期296卷 24-33页

作者： Zhang, Xiaohan Chen, Shaowen Wang, Shuya Huang, Ying Jin, Chuanhong Lin, Fang South China Agr Univ Coll Elect Engn Dept Appl Phys Guangzhou Guangdong Peoples R China Jihua Lab Foshan Guangdong Peoples R China Zhuhai Multiplier Informat Technol Co Ltd Dept Res & Dev Zhuhai Guangdong Peoples R China Zhejiang Univ Sch Mat Sci & Engn State Key Lab Silicon & Adv Semicond Mat Hangzhou Peoples R China South China Agr Univ Coll Elect Engn Guangzhou 510642 Guangdong Peoples R China

High-resolution transmission electron microscopy (HRTEM) images can capture the atomic-resolution details of the dynamically changing structure of nanomaterials. Here, we propose a new scheme and an improved reconstruction algorithm to reconstruct the exit wave function for each image in a focal series of HRTEM images to reveal structural changes. In this scheme, the wave reconstructed from the focal series of images is treated as the initial wave in the reconstruction process for each HRTEM image. Additionally, to suppress noise at the frequencies where the signal is weak due to the modulation of the lens transfer function, a weight factor is introduced in the improved reconstruction algorithm. The advantages of the new scheme and algorithms are validated by using the HRTEM images of a natural specimen and a single-layer molybdenum disulphide. This algorithm enables image resolution enhancement and lens aberration removal, while potentially allowing the visualisation of the structural evolution of nanostructures.

关键词： backpropagation algorithm exit wave function reconstruction high-resolution transmission electron microscopy steepest descent algorithm

来源：评论

学校读者我要写书评

暂无评论

Numerical Simulation and ANN Prediction of Nano-Encapsulated PCM Slurry in a Microchannel: A Thermodynamic Analysis

引用

ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING 2025年 1-21页

作者： Ghasemi, Hamidreza Sheikhzadeh, Ghanbar Ali Fattahi, Abolfazl Univ Kashan Fac Mech Engn Kashan Iran

The growing demand for precise and efficient thermal management in microfluidic heat exchange systems has led to increasing interest in nano-encapsulated phase change materials (NEPCMs) for enhanced heat transfer and thermal energy storage. This study performs a comprehensive numerical simulation and machine learning-based prediction of the thermo-hydrodynamic behavior of NEPCM slurries in microchannels with secondary flow passages. The research aims to quantify the influence of microchannel geometry, flow conditions (Reynolds number: 100-200), and NEPCM concentration (0-10%) on heat transfer and pressure drop characteristics. Energy and entropy analyses are conducted by applying the first and second laws of thermodynamics to assess system efficiency. Furthermore, an artificial neural network (ANN) model is trained to predict the Nusselt number and performance evaluation criterion (PEC) based on input parameters with high accuracy. The simulation results indicate that incorporating NEPCMs enhances heat transfer performance, increasing the average Nusselt number by up to 50% compared to a simple microchannel. However, this improvement comes at the cost of higher pressure drop, with the friction factor showing a variation of up to 100% across different configurations. Entropy generation analysis reveals that thermal entropy generation dominates at lower Reynolds numbers, whereas frictional entropy generation becomes significant at higher Reynolds numbers. The ANN model achieves an R2 value of 0.98, with a prediction error of less than 1.5%, demonstrating its effectiveness. These findings provide quantitative insights for optimizing microchannel-based thermal management systems, balancing heat transfer enhancement, pressure drop, and entropy generation for improved performance in microfluidic applications.

关键词： CFD NEPCM PEC Frictional and thermal entropy generation backpropagation algorithm Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Enhancing Agricultural Disease Detection: A Multi-Model Deep Learning Novel Approach

ENGINEERING REPORTS

引用

ENGINEERING REPORTS 2025年第1期7卷

作者： Hamid, Muhammad Khalid Shah, Said Khalid Husnain, Ghassan Ghadi, Yazeed Yasin Al Maaytah, Shahab Ahmad Qahmash, Ayman Bannu Univ Sci & Technol Dept Comp Sci Bannu Khyber Pakhtunk Pakistan CECOS Univ IT & Emerging Sci Dept Comp Sci Peshawar Pakistan Al Ain Univ Dept Comp Sci & Software Engn Al Ain U Arab Emirates King Faisal Univ Appl Coll Dept Languages & Humanities Alhafof Eastern Provinc Saudi Arabia King Khalid Univ Dept Informat & Comp Syst Abha Saudi Arabia

Artificial intelligence, especially deep learning, has attracted significant interest in bioinformatics, with prominent applications in precision agriculture. A significant threat to the agricultural sector is the rapid propagation of diseases from affected to healthy plants, which, if undetected, may culminate in significant crop losses. This research focusses on employing multi-model deep-learning techniques to identify diseases in the leaves of economically significant crops that are potatoes, tomatoes, grapes, apples, and peaches. These crops are widely grown and crucial for food security, with disease outbreaks threatening yield and quality. This study evaluates the performance of deep learning models, including VGG16, MobileNetV2, Xception, and ResNet, using four metrics, that is, Accuracy, Precision, Recall, and F1-Score. Furthermore, consumer research was undertaken to evaluate user trust in AI-driven multi-model systems, collecting feedback from farmers to inform future research directions. The results demonstrate that the VGG16 model outperformed all others in every evaluation criterion. Experimental simulations were performed in Jupyter Notebook utilizing Anaconda and Python. The findings indicate that the proposed multi-model approach allows a scalable, non-invasive, and contactless machine vision solution for the early detection of diseases in plant leaves, achieving an efficiency of 99% via multimodal classification techniques that incorporate statistical variables including mean, median, mode, skewness, and kurtosis.

关键词： agronomist artificial intelligence backpropagation algorithm convolutional neural network deep learning precision agriculture

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

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

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：