检索结果-内蒙古大学图书馆

Proceedings of the 36th International Conference on Neural Information Processing systems

作者： Markus Hiller Mehrtash Harandi Tom Drummond School of Computing and Information Systems The University of Melbourne Department of Electrical and Computer Systems Engineering Monash University

ISBN: (纸本)9781713871088

Inspired by the concept of preconditioning, we propose a novel method to increase adaptation speed for gradient-based meta-learning methods without incurring extra parameters. We demonstrate that recasting the optimisation problem to a non-linear least-squares formulation provides a principled way to actively enforce a well- conditioned parameter space for meta-learning models based on the concepts of the condition number and local curvature. Our comprehensive evaluations show that the proposed method significantly outperforms its unconstrained counterpart especially during initial adaptation steps, while achieving comparable or better overall results on several few-shot classification tasks - creating the possibility of dynamically choosing the number of adaptation steps at inference time.

关键词：

来源：评论

学校读者我要写书评

暂无评论

ANN-Based Secure Control of Islanded Microgrid Under False Data Injection Cyber-Attack

ANN-Based Secure Control of Islanded Microgrid Under False D...

引用

IEEE Industry Applications Society Annual Meeting (IAS)

作者： Ibtissam Kharchouf Mahmoud S. Abdelrahman Osama A. Mohammed Department of Electrical and Computer Engineering Energy Systems Research Laboratory Florida International University Miami FL USA

AC microgrids may be a target of cyber-attacks such as False Data Injection attack (FDIA) or Denial of Service attacks due to the use of communication technologies. These types of attacks target the communication infrastructure or the control system itself. For instance, FDIAs aim to inject false data into AC microgrids to disrupt the control system. Techniques that detect and mitigate the effect of such attacks are needed. This paper presents an AC microgrid that is structured by parallel inverters and controlled by distributed consensus control strategies to control the frequency and voltage magnitude through active and reactive power. We will investigate the FDIAs effect on the AC microgrid operation. A viable solution to address the FDIA attacks in AC microgrids is developed using an ANN-based reference tracking application to detect and mitigate the attack. The simulation results proved that the proposed strategy could detect and mitigate FDIA attacks especially if the ANN is well trained.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Stochastic Characterization-Based Performance Analysis of an Emulated Communication Network for Cyber- Physical Shipboard Power systems

Stochastic Characterization-Based Performance Analysis of an...

引用

IEEE Electric Ship Technologies Symposium, ESTS

作者： Mahmoud S. Abdelrahman Tung-Lam Nguyen Osama A. Mohammed Department of Electrical and Computer Engineering Energy Systems Research Laboratory Florida International University Miami FL USA

Emerging advanced and innovative Information and Communication Technologies (ICT), automation strategies, and associated algorithms in the traditional power system have transformed it into a Cyber-Physical Power System (CPPS). Energy management and control are crucial in the Shipboard Power System operation. For instance, to meet the ship's energy and other ancillary demands, especially the mission-critical loads that require practical and secure network communication. Therefore, a comprehensive modeling and testing and validation approach was developed in this paper. This enables flexible system-level aspect evaluation through numerous experiments, including virtual, real, and mixed lab settings. This paper presents a scalable and flexible three-layer cyber-physical co-simulation platform implemented in the FIU smart grid testbed. This includes real-time power system simulation using OPAL-RT, ns3 for communication network emulation, and MATLAB/SIMULINK for implementing various applications. To demonstrate the platform's effective operation, energy management with a central controller for SPS was configured and implemented to test the communication network performance by creating tools to characterize the network latency for this application.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Design and Analysis of Voltage Control for Islanded DC Microgrids Based on a Fuzzy-PI Controller

Design and Analysis of Voltage Control for Islanded DC Micro...

引用

IEEE Green Technologies Conference (IEEE-Green)

作者： Hossam Hussein Ahmed Aghmadi Osama A. Mohammed Department of Electrical and Computer Engineering Energy Systems Research Laboratory Florida International University Miami Fl USA

In the transition to clean, cheap, and sustainable energy, microgrid-based renewable energy resources (RES) are widely utilized in islanded or grid-connected modes. However, due to the intermittent nature of RES such as photovoltaic (PV) or wind energy systems, energy storage systems (ESSs) such as batteries are mandated to satisfy load demands and stabilize system operation. For instance, stabilizing the dc bus voltage in islanded microgrids is crucial to keeping the system reliable, dependable, and stable. This paper proposes an adaptive dc bus voltage control technique based on a fuzzy-PI controller. Moreover, a performance comparison between fuzzy-PI and conventional proportional-integrator (PI) controllers is performed based on a MATLAB/Simulink model. The results show that the fuzzy-PI controller has a faster response to any reference voltage variation and less overshoot compared to conventional PI controllers. Additionally, the proposed controller can efficiently stabilize the dc bus voltage during load variation portions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Frequency Stability Enhancement in Low-Inertia Power System Using an Optimal Control Scheme

Frequency Stability Enhancement in Low-Inertia Power System ...

引用

International Conference on Environment and electrical engineering (EEEIC)

作者： Ola Ali Osama A. Mohammed Department of Electrical and Computer Engineering Energy Systems Research Laboratory Florida International University Miami FL USA

Power system dynamics have significantly changed with the global integration of renewable energy sources (RESs). Traditionally, the stored kinetic energy is used for compensating the generation mismatch using the rotating mass of the Synchronous Generators (SGs). With the increase in RESs penetration, SGs are being replaced by inverter-based RESs generation systems which are decoupled from the power systems through power converters. Therefore, the total inertia of the system has decreased, and the system's stability is significantly impacted as the value of the rate of change of frequency (ROCOF) and frequency deviations will be increased. To overcome these challenges, inertia emulation from various energy storage systems (ESS) is used to act as a virtual source of inertia power. This work presents an isolated microgrid (MG) that includes conventional power generation sources and RESs, and an optimal virtual inertia (VI) control loop is proposed for frequency stability enhancement. Supercapacitors (SC) bank with a PID controller presents the VI control loop and acts as a supplementary source for inertia power to support the frequency stability of the system. A genetic algorithm (GA) is used as an optimization technique to get the optimal values of the PID controller gains. The performance of the islanded MG with the proposed optimal control scheme is examined under different operating conditions with varying levels of RESs penetration and system parameters uncertainties. The proposed technique showed its effectiveness and robustness over the system's operation without VI control for keeping the system stable under different operating conditions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Rule-Based Power and Energy Management System For Shipboard Microgrid With HESS To Mitigate Propulsion and Pulsed Load Fluctuations

Rule-Based Power and Energy Management System For Shipboard ...

引用

IEEE Green Technologies Conference (IEEE-Green)

作者： Mahmoud S. Abdelrahman Hossam Hussein Osama A. Mohammed Department of Electrical and Computer Engineering Energy Systems Research Laboratory Florida International University Miami Fl USA

A power and Energy Management System (P/EMS) is crucial to Microgrid's effective practical and reliable operation. In the Shipboard Microgrid (SMG), the load demand is dominated by the high penetration of propulsion loads, which is very dynamic. Some particular loads in naval ships draw very high power for very short periods. A hybrid application of battery and supercapacitor is proposed to cater to the energy consumed by pulse load in a naval ship and for backup purposes. The system proposed in this model is a stand-alone SMG with a Photovoltaic (PV) and Hybrid Energy Storage System (HESS). A rule-based energy supervision and power allocation technique have also been presented considering the specific fuel consumption of diesel engine generators and the characteristics of hybrid energy stored sources (battery and supercapacitor) of the entire SMG. The simulation studies using MATLAB/Simulink software indicate that the proposed P/EMS strategy enables management of power and energy contributions of hybrid resources on the ship. In addition, a reasonable allocation of power among HESS units, which can effectively smooth out the load power fluctuations of the proposed SMG has been introduced.

关键词：

来源：评论

学校读者我要写书评

暂无评论

RE-POSE: Synergizing Reinforcement Learning-Based Partitioning and Offloading for Edge Object Detection

arXiv

引用

arXiv 2025年

作者： Shi, Jianrui Zhao, Yong Cui, Zeyang Shen, Xiaoming Zeng, Minhang Liu, Xiaojie Department of Electrical and Computer Engineering National University of Singapore Singapore Department of Computing The Hong Kong Polytechnic University Hong Kong School of Computing and Data Science The University of Hong Kong Hong Kong Pengcheng Laboratory Shenzhen China

Object detection plays a crucial role in smart video analysis, with applications ranging from autonomous driving and security to smart cities. However, achieving real-time object detection on edge devices presents significant challenges due to their limited computational resources and the high demands of deep neural network (DNN)-based detection models, particularly when processing high-resolution video. Conventional strategies, such as input down-sampling and network up-scaling, often compromise detection accuracy for faster performance or lead to higher inference latency. To address these issues, this paper introduces RE-POSE, a Reinforcement Learning (RL)Driven Partitioning and Edge Offloading framework designed to optimize the accuracy-latency trade-off in resource-constrained edge environments. Our approach features an RL-Based Dynamic Clustering Algorithm (RL-DCA) that partitions video frames into non-uniform blocks based on object distribution and the computational characteristics of DNNs. Furthermore, a parallel edge offloading scheme is implemented to distribute these blocks across multiple edge servers for concurrent processing. Experimental evaluations show that RE-POSE significantly enhances detection accuracy and reduces inference latency, surpassing existing methods. © 2025, CC BY.

关键词： Computation offloading

来源：评论

学校读者我要写书评

暂无评论

ANN-TW Analysis Based Transmission Line Fault Identification and Location with High Penetration of Inverter-Based Resources

ANN-TW Analysis Based Transmission Line Fault Identification...

引用

IEEE Green Technologies Conference (IEEE-Green)

作者： Mahmoud S. Abdelrahman Ibtissam Kharchouf Osama A. Mohammed Department of Electrical and Computer Engineering Energy Systems Research Laboratory Florida International University Miami Fl USA

A Power systems are undergoing a rapid change in generation mix due to the growth of Inverter-Based Resources (IBRs) such as wind and solar PV. This rapid growth creates new challenges for protection engineers. The output current of an IBR facility under short circuit conditions differs significantly from that of a conventional rotating synchronous source facility, posing protection problems. The protection schemes for the transmission line, which were designed locally at the planning stage, may operate reliably for low penetration of IBRs but in the case of high penetration, fault detection and location in the transmission line will be challenged. In this paper, a traveling wave-based protection schemes with the aid of Artificial Neural Network (ANN) as an innovative method to overcome these challenges and facilitate the widespread deployment of IBRs in transmission systems will be presented. A three-bus transmission network with voltage level of 380Kv and 94% of renewables penetration is simulated using DIgSILENT Power Factory is used in this study. The results demonstrates that fault identification is a quick and reliable way to identify a variety of faults, particularly single line to ground faults, which are the most challenging fault for the protection system.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Optimal Modulation and DM Filter Design for a High Switching Frequency Single-Stage Microinverter

Optimal Modulation and DM Filter Design for a High Switching...

引用

Annual IEEE Conference on Applied Power Electronics Conference and Exposition (APEC)

作者： Arafat Hasnain Michael D'Antonio Chanaka Singhabahu Yidi Shen Daniel Zakzewski Alireza Khaligh Department of Electrical and Computer Engineering Maryland Power Electronics Laboratory Institute for Systems Research University of Maryland USA

This paper presents an optimal modulation and systematic filter design approach for a single-stage dual-active-bridge (DAB) based dc-ac microinverter to achieve improved differential-mode (DM) noise performance for electromagnetic interference (EMI) tests. As DM filters contribute significantly to the overall converter volume, the main objective of this work is to leverage the degrees of freedom in the DAB converters to effectively attenuate the EMI noise. In addition, the DM filter design method needs to ensure near unity power factor converter operation. To achieve these targets, this paper analyzes three modulation strategies based on fixed or variable switching frequency operation where the different control modulation variables are varied to find the simulated DM noise spectrum. Based on the required DM attenuation, a constrained optimization problem is formulated to determine minimal DM filter parameters. Simulation results show that a spread spectrum approach with variable switching frequency is shown to minimize the DM EMI attenuation effort by spreading the noise profile. A fully GaN 400 W hardware prototype demonstrating the spread-sprectrum approach.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Transferable Learning of GCN Sampling Graph Data Clusters from Different Power systems

Transferable Learning of GCN Sampling Graph Data Clusters fr...

引用

Annual Allerton Conference on Communication, Control, and computing

作者： Tong Wu Anna Scaglione Daniel Arnold Tianyi Chen The Department of Electrical and Computer Engi-neering University of Central Florida Orlando FL USA The Department of Electrical and Computer Engineering Cornell Tech Cornell University New York City NY USA Lawrence Berkeley National Laboratory The Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute Troy NY USA

ISBN: (数字)9798331541033

ISBN: (纸本)9798331541040

Contemporary neural network (NN) detectors for power systems face two primary challenges. First, each power system requires individual training of NN detectors to accommodate its unique configuration and base demands. Second, significant changes within the power system, such as the introduction of new substations or new generators, necessitate retraining. To overcome these issues, we introduce a novel architecture, the Nodal Graph Convolutional Neural Network (NGCN), which utilizes graph convolutions at each bus and its neighborhoods. This approach allows the training process to encompass multiple power systems and include all buses, thereby enhancing the transferability of the method across different power systems. The NGCN is particularly effective for detection tasks, such as cyber-attacks on smart inverters and false data injection attacks. Our tests demonstrate that the NGCN significantly improves performance over traditional NNs, boosting detection accuracy from approximately 85% to around 97% for the aforementioned task. Furthermore, the transferable NGCN, which is trained by samples from multiple power systems, performs considerably better in evaluations than the NGCN trained on a single power system.

关键词： Training Accuracy Substations Detectors computer architecture Inverters Power systems Cyberattack Faces Monitoring

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：