检索结果-内蒙古大学图书馆

Autonomous Drones in Urban Navigation: autoencoder Learning Fusion for Aerodynamics

JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT 2024年第7期150卷

作者： Wu, Jiahao Ye, Yang Du, Jing Univ Florida Dept Civil & Coastal Engn Informat Cobots & Intelligent Construct Lab Gainesville FL 32611 USA

Drones are becoming indispensable in emergency search and rescue (SAR), particularly in intricate urban areas where rapid and accurate response is crucial. This study addresses the pressing need for enhancing drone navigation in such complex, dynamic urban environments, where obstacles like building layouts and varying wind conditions create unique challenges. Particularly, the need for adapting drone autonomous navigation in correspondence with dynamic wind conditions in urban settings is emphasized because it is important for drones to avoid loss of control or crashes during SAR. This paper introduces a pioneering method integrating multiobjective reinforcement learning (MORL) with a convolutional autoencoder to train autonomous drones in comprehending and reacting to aerodynamic features in urban SAR. MORL enables the drone to optimize multiple goals, whereas the convolutional autoencoder generates synthetic wind simulations with a substantially lower computation cost compared to traditional computational fluid dynamics (CFD) simulations. A unique data transfer structure is also proposed, which fosters a seamless integration of perception and decision-making between machine learning (ML) and reinforcement learning (RL) components. This approach uses imagery data, specific to building layouts, allowing the drone to autonomously formulate policies, prioritize navigation decisions, optimize paths, and mitigate the impact of wind, all while negating the necessity for conventional aerodynamic force sensors. The method was validated with a model of New York City, offering substantial implications for enhancing automation algorithms in urban SAR. This innovation enables the possibility of more efficient, precise, and timely drone SAR operations within intricate urban landscapes.

关键词： Autonomous drones Reinforcement learning (RL) convolutional autoencoder Wind simulation Search and rescue

来源：评论

学校读者我要写书评

暂无评论

Non-intrusive data-driven reduced-order modeling for time-dependent parametrized problems

引用

JOURNAL OF COMPUTATIONAL PHYSICS 2024年 497卷

作者： Duan, Junming Hesthaven, Jan S. Ecole Polytech Fed Lausanne Chair Computat Math & Simulat Sci CH-1015 Lausanne Switzerland Univ Wurzburg Fak Math & Informat Emil Fischer Str 40 D-97070 Wurzburg Germany

Reduced-order models are indispensable for multi-query or real-time problems. However, there are still many challenges to constructing efficient ROMs for time-dependent parametrized problems. Using a linear reduced space is inefficient for time-dependent nonlinear problems, especially for transport-dominated problems. The non-linearity usually needs to be addressed by hyper-reduction techniques, such as DEIM, but it is intrusive and relies on the assumption of affine dependence of parameters. This paper proposes and studies a non-intrusive reduced -order modeling approach for time-dependent parametrized problems. It is purely data-driven and naturally split into offline and online stages. During the offline stage, a convolutional autoencoder, consisting of an encoder and a decoder, is trained to perform dimensionality reduction. The encoder compresses the full-order solution snapshots to a nonlinear manifold or a low-dimensional reduced/latent space. The decoder allows the recovery of the full-order solution from the latent space. To deal with the time-dependent problems, a high-order dynamic mode decomposition (HODMD) is utilized to model the trajectories in the latent space for each parameter. During the online stage, the HODMD models are first utilized to obtain the latent variables at a new time, then interpolation techniques are adopted to recover the latent variables at a new parameter value, and the full-order solution is recovered by the decoder. Some numerical tests are conducted to show that the approach can be used to predict the unseen full-order solution at new times and parameter values fast and accurately, including transport-dominated problems.

关键词： Reduced-order modeling convolutional autoencoder Dynamic mode decomposition Parametrized problem Time-dependent problem Nonlinear problem

来源：评论

学校读者我要写书评

暂无评论

Estimation of the Global Equivalence Ratio of a Swirl Combustor from a Small Number of Absorption Spectra Using Machine Learning

引用

APPLIED SPECTROSCOPY 2024年第10期78卷 1078-1088页

作者： Bong, Cheolwoo Im, Seong-kyun Do, Hyungrok Bak, Moon Soo Sungkyunkwan Univ Sch Mech Engn Suwon South Korea Korea Univ Dept Mech Engn Seoul South Korea Seoul Natl Univ Dept Mech Engn Seoul South Korea Sungkyunkwan Univ Dept Smart Fabricat Technol Suwon South Korea

A new optical diagnostic method that predicts the global fuel-air equivalence ratio of a swirl combustor using absorption spectra from only three optical paths is proposed here. Under normal operation, the global equivalence ratio and total flow rate determine the temperature and concentration fields of the combustor, which subsequently determine the absorption spectra of any combustion species. Therefore, spectra, as the fingerprint for a produced combustion field, were employed to predict the global equivalence ratio, one of the key operational parameters, in this study. Specifically, absorption spectra of water vapor at wavenumbers around 7444.36, 7185.6, and 6805.6 cm-1 measured at three different downstream locations of the combustor were used to predict the global equivalence ratio. As it is difficult to find analytical relationships between the spectra and produced combustion fields, a predictive model was a data-driven acquisition. The absorption spectra as an input were first feature-extracted through stacked convolutional autoencoders and then a dense neural network was used for regression prediction between the feature scores and the global equivalence ratio. The model could predict the equivalence ratio with an absolute error of +/- 0.025 with a probability of 96%, and a gradient-weighted regression activation mapping analysis revealed that the model leverages not only the peak intensities but also the variations in the shape of absorption lines for its predictions. Graphical abstract This is a visual representation of the abstract.

关键词： Swirl combustor status monitoring tunable diode laser absorption spectroscopy TDLAS convolutional autoencoder CAE gradient-weighted regression activation mapping grad-RAM

来源：评论

学校读者我要写书评

暂无评论

A novel self-attention deep subspace clustering

引用

INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS 2021年第8期12卷 2377-2387页

作者： Chen, Zhengfan Ding, Shifei Hou, Haiwei China Univ Min & Technol Sch Comp Sci & Technol Xuzhou 221116 Jiangsu Peoples R China Minist Educ People S Republ China Mine Digitizat Engn Res Ctr Xuzhou 221116 Jiangsu Peoples R China

Most of the existing deep subspace clustering methods leverage convolutional autoencoders to obtain feature representation for non-linear data points. These methods commonly adopt the structure of a few convolutional layers because stacking many convolutional layers may cause computationally inefficient and optimization difficulties. However, long-range dependencies can hardly be captured when convolutional operations are not repeated enough, thus affect the quality of feature extraction which the performance of deep subspace clustering method highly lies in. To deal with this issue, we propose a novel self-attention deep subspace clustering (SADSC) model, which learns more favorable data representations by introducing self-attention mechanisms into convolutional autoencoders. Specifically, SADSC leverages three convolutional layers and add the self-attention layers after the first and third ones in encoders, then decoders have symmetric structures. The self-attention layers maintain the variable input sizes and can be easily combined with different convolutional layers in autoencoder. Experimental results on the handwritten recognition, face and object clustering datasets demonstrate the advantages of SADSC over the state-of-the-art deep subspace clustering models.

关键词： Deep subspace clustering convolutional autoencoder Self-attention

来源：评论

学校读者我要写书评

暂无评论

Multi-objective Genetic Algorithm Based Deep Learning Model for Automated COVID-19 Detection Using Medical Image Data

引用

JOURNAL OF MEDICAL AND BIOLOGICAL ENGINEERING 2021年第5期41卷 678-689页

作者： Bansal, S. Singh, M. Dubey, R. K. Panigrahi, B. K. Indian Inst Technol Delhi Comp Sci & Engn Dept New Delhi 110016 India Robert Bosch Engn & Business Solut Private Ltd He 123 Hosur Rd7th Block Bengaluru 560095 Karnataka India Indian Inst Technol Delhi Elect Engn Dept New Delhi 110016 India

Purpose In early 2020, the world is amid a significant pandemic due to the novel coronavirus disease outbreak, commonly called the COVID-19. Coronavirus is a lung infection disease caused by the Severe Acute Respiratory Syndrome Coronavirus 2 virus (SARS-CoV-2). Because of its high transmission rate, it is crucial to detect cases as soon as possible to effectively control the spread of this pandemic and treat patients in the early stages. RT-PCR-based kits are the current standard kits used for COVID-19 diagnosis, but these tests take much time despite their high precision. A faster automated diagnostic tool is required for the effective screening of COVID-19. Methods In this study, a new semi-supervised feature learning technique is proposed to screen COVID-19 patients using chest CT scans. The model proposed in this study uses a three-step architecture, consisting of a convolutional autoencoder based unsupervised feature extractor, a multi-objective genetic algorithm (MOGA) based feature selector, and a Bagging Ensemble of support vector machines based binary classifier. The proposed architecture has been designed to provide precise and robust diagnostics for binary classification (COVID ***). A dataset of 1252 COVID-19 CT scan images, collected from 60 patients, has been used to train and evaluate the model. Results The best performing classifier within 127 ms per image achieved an accuracy of 98.79%, the precision of 98.47%, area under curve of 0.998, and an F1 score of 98.85% on 497 test images. The proposed model outperforms the current state of the art COVID-19 diagnostic techniques in terms of speed and accuracy. Conclusion The experimental results prove the superiority of the proposed methodology in comparison to existing *** study also comprehensively compares various feature selection techniques and highlights the importance of feature selection in medical image data problems.

关键词： Coronavirus (COVID-19) convolutional autoencoder Multi-objective genetic algorithm Feature subset selection

来源：评论

学校读者我要写书评

暂无评论

A Framework for Anomaly Classification and Segmentation in Remanufacturing using autoencoders and Simulated Data

引用

Procedia CIRP 2024年 126卷 1005-1010页

作者： Jan-Philipp Kaiser Carl-Leandro Enslin Erik Tabuchi Barczak Florian Stamer Michael Heizmann Gisela Lanza wbk Institute of Production Science Karlsruhe Institute of Technology (KIT) Kaiserstr. 12 76131 Karlsruhe IIIT – Institute of Industrial Information Technology Karlsruhe Institute of Technology (KIT) Hertzstr. 16 76187 Karlsruhe Global Advanced Manufacturing Institute (GAMI) KIT China Branch 215123 Suzhou China

In remanufacturing, reusable end-of-life products must be separated from those for which remanufacturing is not economical. In this process, the focus is initially on the detection and evaluation of surface defects of the used product. Due to the use phase of the product, these show a great variety with strongly differing degrees of severity. Deep-learning-based autoencoders offer the advantage of defect-independent detection of anomalies by detecting deviations from a defect-free product condition. In the present work, anomaly detection based on synthetically generated image data is investigated. In addition, a framework for classification and segmentation is developed and presented. The results show that the approach presented in the paper is effective in detecting anomalies with potential for improvement through adjustments such as perceptual loss or image processing methods. However, challenges remain in accurately detecting anomalies in the presence of missing parts, large anomalies, and image data with high variance.

关键词： Remanufacturing Anomaly Detection Unsupervised Learning convolutional autoencoder

来源：评论

学校读者我要写书评

暂无评论

An elevator door anomaly detection method based on improved deep multi-sphere support vector data description

引用

COMPUTERS & ELECTRICAL ENGINEERING 2024年 120卷

作者： Xie, Pengdong Zhang, Linxuan Li, Minghong Qiu, Chaojie Tsinghua Univ Shenzhen Int Grad Sch Shenzhen 518057 Peoples R China Tsinghua Univ Dept Automat Beijing 100084 Peoples R China Xinjiang Univ Sch Elect Engn Urumqi 830017 Peoples R China

Various types of elevator door faults and difficulties in fault data acquisition make it difficult to use supervised learning methods for fault diagnosis. This paper proposes a semi-supervised anomaly detection method based on improved deep multi-sphere support vector data description. Multiple distinguishing hyper-spheres, characterized by minimum volume, are established on the foundation of normal data by this method. These hyper-spheres represent the multi- modal distribution exhibited by the normal data. In addition, the method fuses multi-sensor source data such as tri-axial acceleration, dual-axial tilt angle, and introduces the structure of InceptionTime to realize the fusion of multivariate data and feature extraction in multiple resolutions. Experiments verify the feasibility of the method with an overall AUC of 96.50%, and comparative experiments demonstrate the superior detection performance. This contributes a novel, accurate, and more appropriate method to the elevator door anomaly detection.

关键词： Anomaly detection Support vector data description Semi-supervised learning convolutional autoencoder

来源：评论

学校读者我要写书评

暂无评论

Machine learning-based prediction of swirl combustor operation from flame imaging

引用

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE 2024年 138卷

作者： Bong, Cheolwoo Ali, Mohammed H. A. Im, Seong kyun Do, Hyungrok Bak, Moon Soo Sungkyunkwan Univ Sch Mech Engn Suwon 16419 South Korea Korea Univ Dept Mech Engn Seoul 02841 South Korea Seoul Natl Univ Dept Mech Engn Seoul 08826 South Korea Sungkyunkwan Univ Dept Smart Fab Technol Suwon 16419 South Korea

This study proposes a novel data-driven model to distinguish normal operation of a swirl combustor and predict key operation conditions using a flame image taken with a low-cost monochrome camera. The model, in the form of a convolutional neural network (CNN), is designed to take a flame image as input and provide either the total air flow rate (Q) or the fuel-air equivalence ratio (phi) as an output. However, since the type of problem in this study is regression, it is necessary to make predictions only on normal operation images, as it is not feasible to collect flame images for all unstable combustion modes. Thus, the stacked convolutional layers were first trained as a convolutional autoencoder (CAE) in an unsupervised manner using only flame images under normal operation modes, so that the CAE can perform well only on normal operation images. Then, a regressor that outputs either Q or phi is connected to the trained encoder and trained in a supervised manner. It was found that the model can predict Q and phi within +5.17 L/min (equivalent to 3.4% of the total flow rate) and +0.026, respectively, with a 96% probability, along with detecting abnormalities based on large reconstruction errors of input images. Predictions and image collection can be performed within 50 ms, demonstrating the potential for real-time monitoring of combustor status.

关键词： Swirl combustor Operation condition monitoring Abnormality detection convolutional autoencoder Gradient-weighted activation mapping

来源：评论

学校读者我要写书评

暂无评论

Fault Diagnosis in Centrifugal Pumps: A Dual-Scalogram Approach with Convolution autoencoder and Artificial Neural Network

引用

SENSORS 2024年第3期24卷 851页

作者： Zaman, Wasim Ahmad, Zahoor Kim, Jong-Myon Univ Ulsan Dept Elect Elect & Comp Engn Ulsan 44610 South Korea Prognosis & Diagnost Technol Co Ltd Ulsan 44610 South Korea

This paper proposes a new fault diagnosis method for centrifugal pumps by combining signal processing with deep learning techniques. Centrifugal pumps facilitate fluid transport through the energy generated by the impeller. Throughout the operation, variations in the fluid pressure at the pump's inlet may impact the generalization of traditional machine learning models trained on raw statistical features. To address this concern, first, vibration signals are collected from centrifugal pumps, followed by the application of a lowpass filter to isolate frequencies indicative of faults. These signals are then subjected to a continuous wavelet transform and Stockwell transform, generating two distinct time-frequency scalograms. The Sobel filter is employed to further highlight essential features within these scalograms. For feature extraction, this approach employs two parallel convolutional autoencoders, each tailored for a specific scalogram type. Subsequently, extracted features are merged into a unified feature pool, which forms the basis for training a two-layer artificial neural network, with the aim of achieving accurate fault classification. The proposed method is validated using three distinct datasets obtained from the centrifugal pump under varying inlet fluid pressures. The results demonstrate classification accuracies of 100%, 99.2%, and 98.8% for each dataset, surpassing the accuracies achieved by the reference comparison methods.

关键词： centrifugal pump Stockwell transform fault diagnosis rotating machinery convolutional autoencoder vibrational signals deep learning

来源：评论

学校读者我要写书评

暂无评论

RL Based Unsupervised Video Summarization Framework for Ultrasound Imaging 3rd

RL Based Unsupervised Video Summarization Framework for Ultr...

引用

3rd International Workshop of Advances in Simplifying Medical Ultrasound (ASMUS)

作者： Mathews, Roshan P. Panicker, Mahesh Raveendranatha Hareendranathan, Abhilash R. Chen, Yale Tung Jaremko, Jacob L. Buchanan, Brian Narayan, Kiran Vishnu Chandrasekharan, Kesavadas Mathews, Greeta Indian Inst Technol Palakkad Palakkad India Univ Alberta Edmonton AB Canada Hosp Univ Puerta de Hierro Spain Madrid Spain Govt Med Coll Thiruvananthapuram Thiruvananthapuram Kerala India Sree Chitra Inst Med Sci & Technol Thiruvananthap Thiruvananthapuram Kerala India Bhagwan Mahaveer Jain Hosp Bangalore Bangalore Karnataka India

ISBN: (纸本)9783031169021;9783031169014

The need for summarizing long medical scan videos for automatic triage in Emergency Departments and transmission of the summarized videos for telemedicine has gained significance during the COVID-19 pandemic. However, supervised learning schemes for summarizing videos are infeasible as manual labeling of scans for large datasets is impractical by frontline clinicians. This work presents a methodology to summarize ultrasound videos using completely unsupervised learning schemes and is validated on Lung Ultrasound videos. A convolutional autoencoder and a Transformer decoder is trained in an unsupervised reinforcement learning setup i.e., without supervised labels in the whole workflow. Novel precision and recall computation for ultrasound videos is also presented employing which high Precision and F 1 scores of 64.36% and 35.87% with an average video compression rate of 78% is obtained when validated against clinically annotated cases. Even though demonstrated using lung ultrasound videos, our approach can be readily extended to other imaging modalities.

关键词： Ultrasound Video summarization Unsupervised reinforcement learning convolutional autoencoder Transformer

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：