检索结果-内蒙古大学图书馆

On the vanishing and exploding gradient problem in Gated Recurrent Units

IFAC-PapersOnLine 2020年第2期53卷 1243-1248页

作者： Alexander Rehmer Andreas Kroll Department of Measurement and Control Institute for System Analytics and Control Faculty of Mechanical Engineering University of Kassel Germany

Recurrent Neural Networks are applied in areas such as speech recognition, natural language and video processing, and the identification of nonlinear state space models. Conventional Recurrent Neural Networks, e.g. the Elman Network, are hard to train. A more recently developed class of recurrent neural networks, so-called Gated Units, outperform their counterparts on virtually every task. This paper aims to provide additional insights into the differences between RNNs and Gated Units in order to explain the superior perfomance of gated recurrent units. It is argued, that Gated Units are easier to optimize not because they solve the vanishing gradient problem, but because they circumvent the emergence of large local gradients.

关键词： Nonlinear system identification Recurrent Neural Networks Gated Recurrent Units

来源：评论

学校读者我要写书评

暂无评论

Deep and Machine Learning-based Methods for Defect Classification in Jet Engines

Deep and Machine Learning-based Methods for Defect Classific...

引用

Intermountain engineering, Technology and Computing (IETC)

作者： Marco P. Schoen Marcel Oettinger Dajan Mimic Department of Mechanical Engineering Measurement and Control Engineering Research Center Idaho State University Pocatello USA MTU Aero Engines AG Formerly Institute of Turbomachinery and Fluid Dynamics Munich Germany Institute of Turbomachinery and Fluid Dynamics Leibniz Universität Hannover Hannover Germany Cluster of Excellence SE2A - Sustainable and Energy Efficient Aviation Technische Universität Braunschweig Braunschweig Germany

In this paper, the utility and accuracy of Machine Learning (ML) and Deep Learning (DL) methods are investigated for detecting defects in civil aircraft engines. Rather than to disassemble jet engines, the approach investigated in this study utilizes images of the exhaust of jet engines and infers defects in the turbine and burner section. While the proposed DL methods make use of one or two cameras, the ML methods depend on data obtained by extracting the density fields of the Hot Gas Path (HGP). The HPG data are computed from images acquired by an array of cameras. The corresponding ML features are crafted from these density fields. The proposed algorithms employ optimized hyperparameters and separate training as well as validation data sets. The study illustrates the potential of DL methods and the resulting simplification in the necessary instrumentation to accomplish near perfect defect classification outcomes.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Numerical Investigation of Indoor Air Quality in a Fitting Room Equipped with the Mixing and Displacement Ventilation Systems

SSRN

引用

SSRN 2023年

作者： Bennia, Abderazak Khan, Mohammed Abdul Hameed Lateb, Mohamed Fellouah, Hachimi Laboratory of Control Testing Measurement and Simulation in Mechanics Université Hassiba Benbouali Chlef02000 Algeria Department of Electromechanics Université Mohamed El-Bachir El-Ibrahimi Bordj Bou Arreridj34030 Algeria Department of Mechanical Engineering Université de Sherbrooke SherbrookeQCJ1K 2R1 Canada

In this research, computational fluid dynamics is used to examine the indoor air quality in a fitting room incorporated with two mechanical ventilation systems, namely mixing ventilation (MV) and displacement ventilation (DV) *** two ventilation systems capacity to remove the airborne pollutant from the room is measured using the mean age of air technique. Air quality indicators, such as ventilation efficiency and air change effectiveness, are also used to assess the quality of the air in the space of the *** findings showed that the placement of the inlet and outlet, in addition to the kind of mechanical ventilation system, affect the delivery of fresh air to the occupants and either attenuate or intensify the spread of pollutants in the room space. When compared to the MV system, the evaluation showed that the DV system performs successfully in terms of the system capacity to distribute air and remove internally produced pollutants. In fact, the benchmarking showed that the DV system works well since it offers little contamination dispersion in the room space and exhibitsa better quantity/quality of fresh air at the breathing level © 2023, The Authors. All rights reserved.

关键词： Air quality

来源：评论

学校读者我要写书评

暂无评论

Second harmonic generation of laser beams in transverse mode locking states

引用

Advanced Photonics 2022年第2期4卷 68-77页

作者： Zilong Zhang Yuan Gao Xiangjia Li Xin Wang Suyi Zhao Qiang Liu Changming Zhao Beijing Institute of Technology School of Optics and PhotonicsBeijingChina Ministry of Education Key Laboratory of Photoelectronic Imaging Technology and SystemBeijingChina Ministry of Industry and Information Technology Key Laboratory of Photonics Information TechnologyBeijingChina Arizona State University School for Engineering of MatterTransport and EnergyDepartment of Aerospace and Mechanical EngineeringTempeArizonaUnited States Ministry of Education Key Laboratory of Photonic Control Technology(Tsinghua University)BeijingChina Tsinghua University Department of Precision InstrumentState Key Laboratory of Precision Measurement Technology and InstrumentsBeijingChina

Nonlinear frequency conversion of structured beams has been of great interest *** present an intracavity second harmonic generation(SHG)of laser beams in transverse mode locking(TML)states with a specially designed sandwich such as a microchip *** intracavity nonlinear frequency conversion process of a laser beam in a TML state to its second harmonic is theoretically and experimentally investigated,considering different relative phase and weight parameters between the basic modes in the TML *** between the far-field SHG beam patterns of fundamental frequency transverse modes in coherently locked and incoherently superposed states demonstrates that the SHG of TML beams can carry more *** rarely observed far-field SHG beam patterns are obtained,and they are consistent with the theoretical analysis and numerical *** the obtained SHG beams,the characteristics of the structured fundamental frequency beams can also be conversely investigated or *** work may have important applications in optical 3D printing,optical trapping of particles,and free-space optical communication areas.

关键词： transverse mode locking optical vortex lattice structured beams second harmonic generation

来源：评论

学校读者我要写书评

暂无评论

Optimization of Directional Landmark Deployment for Visual Observer on SE(3)

arXiv

引用

arXiv 2022年

作者： Lei, Zike Chen, Xi Tan, Ying Chen, Xiang Chai, Li The School of Information Science and Engineering Wuhan University of Science and Technology Hubei 430081 China The Engineering Research Center of Metallurgical Automation and Measurement Technology Wuhan University of Science and Technology Hubei 430081 China The School of Electrical Mechanical and Infrastructure Engineering The University of Melbourne VIC3010 Australia The Department of Electrical and Computer Engineering University of Windsor ONN9B 3P4 Canada The College of Control Science and Engineering Zhejiang University Hangzhou310027 China

An optimization method is proposed in this paper for novel deployment of given number of directional landmarks (location and pose) within a given region in the 3-D task space. This new deployment technique is built on the geometric models of both landmarks and the monocular camera. In particular, a new concept of Multiple Coverage Probability (MCP) is defined to characterize the probability of at least n landmarks being covered simultaneously by a camera at a fixed position. The optimization is conducted with respect to the position and pose of the given number of landmarks to maximize MCP through globally exploration of the given 3-D space. By adopting the elimination genetic algorithm, the global optimal solutions can be obtained, which are then applied to improve the convergent performance of the visual observer on SE(3) as a demonstration example. Both simulation and experimental results are presented to validate the effectiveness of the proposed landmark deployment optimization method. © 2022, CC BY-NC-SA.

关键词： Probability

来源：评论

学校读者我要写书评

暂无评论

Numerical Investigation of Indoor Air Quality in a Fitting Room Equipped with the Mixing and Displacement Ventilation Systems

SSRN

引用

SSRN 2022年

作者： Bennia, Abderazak Hameed Khan, Mohammed Abdul Lateb, Mohamed Fellouah, Hachimi Laboratory of Control Testing Measurement and Simulation in Mechanics Université Hassiba Benbouali Chlef02000 Algeria Department of Electromechanics Université Mohamed El-Bachir El-Ibrahimi Bordj Bou Arreridj34030 Algeria Department of Mechanical Engineering Université de Sherbrooke SherbrookeQCJ1K 2R1 Canada

In this research, computational fluid dynamics is used to examine the indoor air quality in a fitting room incorporated with two mechanical ventilation systems, namely mixing ventilation (MV) and displacement ventilation (DV) systems. The two ventilation systems capacity to remove the airborne pollutant from the room is measured using the mean age of air technique. Air quality indicators, such as ventilation efficiency and air change effectiveness, are also used to assess the quality of the air in the space of the room. The findings showed that the placement of the inlet and outlet, in addition to the kind of mechanical ventilation system, affect the delivery of fresh air to the occupants and either attenuate or intensify the spread of pollutants in the room space. When compared to the MV system, the evaluation showed that the DV system performs successfully in terms of the system capacity to distribute air and remove internally produced pollutants. In fact, the benchmarking showed that the DV system works well since it offers little contamination dispersion in the room space and exhibits a better quantity/quality of fresh air at the breathing level. © 2022, The Authors. All rights reserved.

关键词： Computational fluid dynamics

来源：评论

学校读者我要写书评

暂无评论

Numerical investigation on aerodynamic characteristics of two rotating cylinders in tandem

引用

Physics of Fluids 2025年第6期37卷

作者： Wang, Jialin Xue, Xiaopeng School of Automation Central South University Changsha410083 China Department of Mechanical and Mechatronics Engineering Waterloo University of Waterloo Waterloo Canada Hunan Provincial Key Laboratory of Optic-Electronic Intelligent Measurement and Control Changsha410083 China

To solve the problem of unsteady wake and high drag in the flow around a cylinder, combined with the advantages of two cylinders in tandem in drag reduction and rotating cylinder in restraining wake oscillation, the k − ω Shear Stress Transport turbulence model was used to explore the aerodynamic characteristics and flow field structures of two rotating cylinders in tandem at different speed ratios under the condition of Re = 200, gap ratio L/D = 1.5, 2.5, and 3.5, respectively. The effectiveness of the method is verified by comparing the results with the available experiments. The streamline, periodic averaged lift and drag coefficient and the pressure coefficient on the cylinder surface are given and analyzed in detail. It was found that for the rotating single cylinder, when the speed ratio is 4.1, it has best the aerodynamic performance. For the rotating double cylinders, the higher gap ratio can increase the combined lift coefficient and reduce the combined drag coefficient under the condition of bigger speed ratios of cylinder 1. When the gap ratio is 3.5, the rear cylinder is stationary or has a higher speed ratio, it can restrain the wake oscillation of the front cylinder. When the front to rear cylinder speed ratio is 2 and 1, respectively, the aerodynamic characteristics of the rotating double cylinder system have best performances. © 2025 Author(s).

关键词： Shear stress

来源：评论

学校读者我要写书评

暂无评论

Federated Multi-Agent Deep Reinforcement Learning Approach via Physics-Informed Reward for Multi-Microgrid Energy Management

arXiv

引用

arXiv 2022年

作者： Li, Yuanzheng He, Shangyang Li, Yang Shi, Yang Zeng, Zhigang School of Artificial Intelligence and Automation Key Laboratory on Image Information Processing and Intelligent Control of Ministry of Education Huazhong University of Science and Technology Wuhan430074 China China-Belt and Road Joint Laboratory on Measurement and Control Technology Wuhan430074 China China-EU Institute for Clean and Renewable Energy Huazhong University of Science and Technology Wuhan430074 China School of Electrical Engineering Northeast Electric Power University Jilin132012 China The Department of Mechanical Engineering University of Victoria VictoriaBC V8P 5C2 Canada

The utilization of large-scale distributed renewable energy promotes the development of the multi-microgrid (MMG), which raises the need of developing an effective energy management method to minimize economic costs and keep self energy-sufficiency. The multi-agent deep reinforcement learning (MADRL) has been widely used for the energy management problem because of its real-time scheduling ability. However, its training requires massive energy operation data of microgrids (MGs), while gathering these data from different MGs would threaten their privacy and data security. Therefore, this paper tackles this practical yet challenging issue by proposing a federated multi-agent deep reinforcement learning (F-MADRL) algorithm via the physics-informed reward. In this algorithm, the federated learning (FL) mechanism is introduced to train the F-MADRL algorithm thus ensures the privacy and the security of data. In addition, a decentralized MMG model is built, and the energy of each participated MG is managed by an agent, which aims to minimize economic costs and keep self energy-sufficiency according to the physics-informed reward. At first, MGs individually execute the self-training based on local energy operation data to train their local agent models. Then, these local models are periodically uploaded to a server and their parameters are aggregated to build a global agent, which will be broadcasted to MGs and replace their local agents. In this way, the experience of each MG agent can be shared and the energy operation data is not explicitly transmitted, thus protecting the privacy and ensuring data security. Finally, experiments are conducted on Oak Ridge national laboratory distributed energy control communication lab microgrid (ORNL-MG) test system, and the comparisons are carried out to verify the effectiveness of introducing the FL mechanism and the outperformance of our proposed F-MADRL. Copyright © 2022, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Black-box domain adaptation for cross-domain on-device machinery fault diagnosis via hierarchical debiased self-supervised learning

引用

Knowledge-Based Systems 2025年 324卷

作者： Zhu, Mengliang Liu, Jie Lu, Yanglong Hu, Zhongxu Yuan, Li Zhou, Kaibo MOE Key Laboratory of Image Information Processing and Intelligent Control Belt and Road Joint Laboratory on Measurement and Control Technology School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China National Center of Technology Innovation for Digital Construction School of Civil and Hydraulic Engineering Huazhong University of Science and Technology Wuhan430074 China Department of Mechanical and Aerospace Engineering The Hong Kong University of Science and Technology Clear Water Bay Hong Kong SAR China School of Mechanical Science and Engineering Huazhong University of Science and Technology Wuhan430074 China

This paper explores black-box domain adaptation (BBDA) for cross-domain on-device machinery fault diagnosis. Specifically, a pre-trained black-box source model is deployed on a cloud platform with only its input-output API accessible, and a randomly initialized target model is trained locally utilizing unlabeled target data on an edge device. In BBDA, neither raw data nor model parameters are transmitted across domains, and the only available supervised information for adaptation is API-queried predictions on unlabeled target data. However, API-queried predictions are inevitably noisy and class-imbalanced due to cross-domain distribution discrepancy and inherent category bias. Directly utilizing them for on-device training easily leads to confirmation bias and exacerbates category bias, thereby severely degrading performance. This paper proposes a hierarchical debiased self-supervised learning framework for BBDA from a novel research perspective of concurrently alleviating both confirmation and category biases. Specifically, debiased knowledge distillation is proposed to gradually enhance discriminability of the randomly initialized target model by providing debiased training predictions. An adaptive category-unbiased sample division strategy is devised to divide unlabeled target data into certain-aware and uncertain-aware sets, enabling fully exploiting intrinsic data structures and utilizing underlying sample relationships within the target domain. On this basis, debiased prototypical self-training is proposed for the certain-aware set to obtain reliable pseudo-labels for self-training, leveraging global semantic structure. Hierarchical neighborhood adaptation is customized for the uncertain-aware set to enforce the cluster assumption for noise refinement, exploiting local clustering structure and utilizing underlying sample relationships. Extensive experiments and analyses on public and practical datasets validate the proposed framework's superiority in cross-dom

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Parallel Planning:A New Motion Planning Framework for Autonomous Driving

引用

IEEE/CAA Journal of Automatica Sinica 2019年第1期6卷 236-246页

作者： Long Chen Xuemin Hu Wei Tian Hong Wang Dongpu Cao Fei-Yue Wang IEEE School of Data and Computer Science Sun Yat-sen University the School of Computer Science and Information Engineering Hubei University Institute of Measurement and Control Systems Karlsruhe Institute of Technology Department of Mechanical and Mechatronics Engineering University of Waterloo the State Key Laboratory of Management and Control for Complex Systems.Institute of Automation Chinese Academy of Sciences the Research Center for Military Computational Experiments and Parallel Systems Technology National University of Defense Technology

Motion planning is one of the most significant technologies for autonomous driving. To make motion planning models able to learn from the environment and to deal with emergency situations, a new motion planning framework called as"parallel planning" is proposed in this paper. In order to generate sufficient and various training samples, artificial traffic scenes are firstly constructed based on the knowledge from the reality.A deep planning model which combines a convolutional neural network(CNN) with the Long Short-Term Memory module(LSTM) is developed to make planning decisions in an end-toend mode. This model can learn from both real and artificial traffic scenes and imitate the driving style of human ***, a parallel deep reinforcement learning approach is also presented to improve the robustness of planning model and reduce the error rate. To handle emergency situations, a hybrid generative model including a variational auto-encoder(VAE) and a generative adversarial network(GAN) is utilized to learn from virtual emergencies generated in artificial traffic scenes. While an autonomous vehicle is moving, the hybrid generative model generates multiple video clips in parallel, which correspond to different potential emergency scenarios. Simultaneously, the deep planning model makes planning decisions for both virtual and current real scenes. The final planning decision is determined by analysis of real observations. Leveraging the parallel planning approach, the planner is able to make rational decisions without heavy calculation burden when an emergency occurs.

关键词： Autonomous driving artificial traffic scene deep learning emergencies motion planning parallel planning

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：