Article
Telecommunications
Mounssif Krouka, Anis Elgabli, Chaouki Ben Issaid, Mehdi Bennis
Summary: This paper proposes a novel distributed reinforcement learning algorithm, A-RLADMM, that tackles the challenge of wireless channel interference using the alternating direction method of multipliers (ADMM) and analog transmission scheme.
IEEE TRANSACTIONS ON COGNITIVE COMMUNICATIONS AND NETWORKING
(2022)
Article
Thermodynamics
Simone Totaro, Ioannis Boukas, Anders Jonsson, Bertrand Cornelusse
Summary: This paper addresses the lifelong control problem of isolated microgrids, proposing a novel model-based reinforcement learning algorithm that can handle system changes and outperforms traditional algorithms.
Article
Energy & Fuels
Jian Chen, Khalid Alnowibet, Andres Annuk, Mohamed A. Mohamed
Summary: This article proposes an effective energy management framework for a distributed networked microgrid and enhances the accuracy of modeling parameter uncertainties using a reinforcement learning approach. Additionally, a secure energy transaction structure based on the directed acyclic graph method is introduced to establish a secure communication platform.
ENERGY STRATEGY REVIEWS
(2021)
Article
Multidisciplinary Sciences
J. Aznar-Poveda, A-J Garcia-Sanchez, E. Egea-Lopez, J. Garcia-Haro
Summary: This study addresses the issue of channel load in vehicular communications by utilizing non-cooperative, distributed beaconing allocation. By formulating the problem as a Markov Decision Process and solving it using approximate reinforcement learning, the study achieves faster delivery of emergency-related notifications and good overall performance compared to traditional solutions.
SCIENTIFIC REPORTS
(2022)
Article
Energy & Fuels
Yijun Cheng, Jun Peng, Kaiyang Liu, Fu Jiang, Yue Wu, Zhiwu Huang
Summary: This study proposes a new distributed energy management scheme that combines ADMM and symmetric ADMM algorithms, optimizes decisions using data from the previous day, and achieves the goals of reducing system operation costs and maintaining battery status.
SUSTAINABLE ENERGY GRIDS & NETWORKS
(2022)
Article
Computer Science, Information Systems
Jiaju Qi, Lei Lei, Kan Zheng, Simon X. Yang, Xuemin Shen
Summary: In this article, the scheduling issue of diesel generators (DGs) in an Internet of Things (IoT)-Driven isolated microgrid (MG) is investigated using deep reinforcement learning (DRL). A novel finite-horizon partial observable Markov decision process (POMDP) model is conceived to handle the challenge of discrete-continuous hybrid action space. The proposed DRL algorithm, HAFH-RDPG, seamlessly integrates two classical DRL algorithms and is evaluated using real-world data in an IoT-driven MG.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Shuhua Gao, Cheng Xiang, Ming Yu, Kuan Tak Tan, Tong Heng Lee
Summary: This paper investigates the economic operation of a microgrid with various distributed energy resources, and proposes a novel data-driven online approach based on imitation learning. By decomposing the problem and mimicking a mixed-integer linear programming solver, this method overcomes the limitations of reinforcement learning methods and achieves significant training time reduction and operation cost optimization.
IEEE TRANSACTIONS ON SMART GRID
(2022)
Article
Thermodynamics
Chenyu Guo, Xin Wang, Yihui Zheng, Feng Zhang
Summary: The paper introduces a real-time dynamic optimal energy management method based on deep reinforcement learning algorithm to address energy management issues in microgrids effectively. Compared to traditional methods, the proposed approach offers higher computational accuracy and efficiency. Through offline training and online operation, the algorithm can learn from historical data to capture the uncertainty characteristics of renewable energy and load consumption.
Article
Mathematics
Bowen Zhou, Zhibo Zhang, Chao Xi, Boyu Liu
Summary: This paper proposes a novel two-stage, dual-layer distributed optimization operational approach for microgrids with EVs, which ensures consistent charging/discharging and SOC control of EVs through the lower distributed control layer and determines the optimal operational strategy of the microgrid through the upper optimization scheduling layer.
Article
Computer Science, Information Systems
Yang Xiao, Jun Liu, Jiawei Wu, Nirwan Ansari
Summary: After decades of unprecedented development, modern networks have surpassed expectations in scale and complexity, traditional traffic engineering approaches are inadequate for meeting the quality of service requirements. Deep reinforcement learning has shown to be a feasible solution for autonomously controlling and managing complex systems, with applications benefiting the communications industry.
IEEE COMMUNICATIONS SURVEYS AND TUTORIALS
(2021)
Article
Computer Science, Hardware & Architecture
Huihui Ma, Du Xu, Yueyue Dai, Qing Dong
Summary: This paper introduces the use of Deep Reinforcement Learning to study the AQM scheme, proposing the DRL-AQM model that performs well in various complex network scenarios and can adapt to the dynamic changes in the network.
Article
Energy & Fuels
Imen Jendoubi, Francois Bouffard
Summary: This study proposes a novel RL-based energy management approach for scheduling controllable devices within an electric network. It efficiently solves multi-dimensional, multi-objective, and partially observable power system problems. The novelty lies in the implementation of a hierarchical RL-based control strategy, the introduction of multi-agent reinforcement learning (MARL) for efficient coordination, and the integration of hierarchical RL and MARL theory for a robust control framework. Comparative performance evaluation and experimental results validate the effectiveness of the proposed control framework.
Article
Computer Science, Artificial Intelligence
Gang Bao, Rui Xu
Summary: In this paper, a data-driven energy management strategy based on prioritized experience replay soft actor-critic (PERSAC) is proposed for microgrid (MG) systems. The algorithm optimizes the operation cost by controlling distributed generators (DGs) and an energy storage system (ESS). Numerical examples show that the proposed algorithm achieves the lowest operation cost compared to other algorithms.
COGNITIVE COMPUTATION
(2023)
Article
Engineering, Multidisciplinary
Mohammed H. Alabdullah, Mohammad A. Abido
Summary: This paper proposes a deep reinforcement learning-based approach to optimally manage different energy resources within a microgrid. The proposed methodology considers the stochastic behavior of the main elements and formulates the energy management problem as a finite horizon Markov Decision Process (MDP) without prior knowledge of transition probabilities. An efficient reinforcement learning algorithm based on deep Q-networks is implemented to solve the problem. The results of a case study based on a real microgrid demonstrate the effectiveness of the proposed methodology in achieving optimal cost-effective scheduling of energy resources under stochastic conditions.
ALEXANDRIA ENGINEERING JOURNAL
(2022)
Article
Chemistry, Multidisciplinary
Guodong Guo, Yanfeng Gong
Summary: This paper proposes a multi-agent reinforcement learning algorithm for real-time energy management of a multi-microgrid (MMG) system. The MMG is connected to a distribution network, and the distribution network operator and each microgrid are modeled as autonomous agents. Each agent makes decisions based on local information, and the problem is modeled as a Markov game and solved using the prioritized multi-agent deep deterministic policy gradient (PMADDPG) algorithm.
APPLIED SCIENCES-BASEL
(2023)
Article
Energy & Fuels
Mehdi Montakhabi, Fairouz Zobiri, Shenja van der Graaf, Geert Deconinck, Domenico Orlando, Pieter Ballon, Mustafa A. Mustafa
Summary: This article introduces new roles in future peer-to-peer electricity trading markets through analyzing the value network of the current electricity market and future scenarios based on user involvement and customer ownership. The study assists policy makers in considering future requirements and prompts current market players to reconsider their business models.
Article
Construction & Building Technology
Christophe Patyn, Geert Deconinck
Summary: This paper introduces a data-driven methodology for modeling and controlling residential heating systems and buildings. Dynamic mode decomposition is utilized to fit a linear state-space model, resulting in prediction accuracy similar to greybox models. To address weather prediction uncertainty, a robust linear model predictive control framework is used.
ENERGY AND BUILDINGS
(2022)
Article
Engineering, Electrical & Electronic
Iason I. Avramidis, Florin Capitanescu, Geert Deconinck
Summary: This paper introduces an extensive multi-period optimal power flow framework for smart LV distribution systems, focusing on residential flexibility to combat operational issues. The study includes detailed performance assessments on ZIP flexible loads, controllability of conventional residential devices, and exploitation of inverter technologies. The analysis highlights the de-congestion potential of common residential devices and explores setups with partial adjustments on customer profiles.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2021)
Article
Energy & Fuels
Rohit Trivedi, Sandipan Patra, Yousra Sidqi, Benjamin Bowler, Fiona Zimmermann, Geert Deconinck, Antonios Papaemmanouil, Shafi Khadem
Summary: This article discusses suitable approaches for empowering energy citizens and smart energy communities through community-based microgrid solutions. It addresses functional architectural layers, system integration topologies, interoperability issues, strategies for consumer-centric energy trading, and socio-economic aspects. The article also reviews the concept of local electricity market (LEM) and peer-to-peer (P2P) energy exchange, and explores the development of C-MG and energy trading facilities with interoperability issues. The overall contribution of the article is to pave the path for advanced research and industrialization in the field of smart energy communities.
Article
Engineering, Electrical & Electronic
Iason-Iraklis Avramidis, Florin Capitanescu, Vasileios A. Evangelopoulos, Pavlos S. Georgilakis, Geert Deconinck
Summary: Following EU directives, the adoption of smart sustainable buildings in low voltage networks is expected to increase. This study proposes a coordination framework driven by smart sustainable buildings for the provision of ancillary services. The framework utilizes a novel model predictive control approach and considers collaboration between the distribution system operator and smart sustainable buildings.
IEEE TRANSACTIONS ON SMART GRID
(2022)
Article
Engineering, Electrical & Electronic
Boyuan Wei, Fairouz Zobiri, Geert Deconinck
Summary: This paper proposes a real-time peer to peer flexibility trading scheme for low voltage distribution networks (LVDNs) to address users not always following regulations. The scheme utilizes player compatibility equilibrium and Bernoulli trials for flexibility allocation and consensus and voltage reference for real-time flexibility trading. The proposed scheme is computationally tractable and can protect user privacy. Case studies demonstrate the effectiveness and robustness of the scheme.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2022)
Article
Green & Sustainable Science & Technology
Iason Avramidis, Florin Capitanescu, Geert Deconinck
Summary: This paper presents a method to analyze and monetize the flexibility of computer grid, with a focus on grid-friendly smart sustainable buildings. The impact of different types of device power factor on pricing and envelope reporting is explored, and the method is validated through a generic case study.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY
(2022)
Article
Energy & Fuels
Emilio J. Palacios-Garcia, Xavier Carpent, Joppe W. Bos, Geert Deconinck
Summary: Residential demand side management (DMS) is an effective tool for maintaining network balance, but privacy concerns related to user personal consumption data have hindered its widespread adoption. This paper proposes a privacy-preserving aggregation algorithm based on additive random shares and a combination of symmetric and asymmetric key cryptography methods. The algorithm is compared with other techniques such as additive homomorphic encryption (AHE) and state-of-the-art MPC protocols. The results show that while generic techniques like homomorphic encryption are computationally expensive, MPC approaches provide better performance and resilience for large networks. The proposed additive random shares algorithm is considered the most balanced choice for DSM, offering good performance, simpler information flow, and the ability to add redundant intermediary parties for enhanced resilience.
Article
Engineering, Electrical & Electronic
Sander Claeys, Frederik Geth, Geert Deconinck
Summary: This paper investigates the optimal power flow problem in low-voltage distribution networks and provides formulations for four-wire networks in current-voltage and power-voltage variables, comparing their robustness and scalability.
ELECTRIC POWER SYSTEMS RESEARCH
(2022)
Article
Engineering, Electrical & Electronic
Iason-Iraklis Avramidis, Florin Capitanescu, Geert Deconinck
Summary: Modern distribution networks face frequent operational issues due to the abundance of distributed energy resources (DERs) and outdated infrastructure and management mentality. Distribution system operators (DSOs) must decide whether to reinforce or utilize local flexibility markets to solve these problems, considering the cost of flexibility and value of deferred grid reinforcement. This study proposes a 3-step methodology for defining network-specific, fixed flexibility price catalogues for different DERs, aiming to protect the financial interests of DSOs and end-users and establish a fair value for flexibility.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2023)
Article
Engineering, Electrical & Electronic
Iason-Iraklis Avramidis, Florin Capitanescu, Geert Deconinck, Himanshu Nagpal, Per Heiselberg, Andre Madureira
Summary: Recent geopolitical developments have made global decarbonization a top priority, but transitioning to a low-carbon energy system is challenging. The rapid growth of renewable technologies introduces unprecedented uncertainty in decision-making for operators, and increasing electrification strains the power grid. Electricity customers now play an active role in the system rather than being passive, and demand-side flexibility is seen as crucial to stimulate customers to modify their consumption patterns and contribute to energy supply through local production. Residential, office, and commercial buildings, which account for about 30%-40% of total energy consumption, offer significant potential for providing flexibility services.
IEEE POWER & ENERGY MAGAZINE
(2023)
Proceedings Paper
Computer Science, Theory & Methods
Shicong Zhang, Klaas Thoelen, Thijs Peirelinck, Geert Deconinck
Summary: As the EV market grows, smart charging is crucial to maximize users' benefits and mitigate the impact on the distribution network. This study applies data-driven methods to forecast the energy demand of EVs in an office building case in Belgium. The findings provide valuable insights into charging schedule planning for future smart charging services.
2023 IEEE BELGRADE POWERTECH
(2023)
Proceedings Paper
Computer Science, Theory & Methods
Chunlei Ge, Runze Yu, Iason I. Avramidis, Geert Deconinck
Summary: The concept of smart sustainable buildings is crucial for developing smart cities. This paper proposes a data-driven local control scheme for controllable devices in these buildings, using machine learning algorithms and historical data. The scheme improves the safety of distribution networks with a higher penetration rate of controllable devices.
2023 IEEE BELGRADE POWERTECH
(2023)
Proceedings Paper
Computer Science, Theory & Methods
M. Furqan Azam, Thijs Becker, Chris Hermans, Koen Vanthournout, Geert Deconinck
Summary: Due to the increased penetration of low-carbon technologies, low voltage distribution networks are expected to face congestion and voltage violation issues. This paper proposes a novel scenario generation technique to accurately sample optimal load profiles for LV customers. The technique combines clustering and weight learning strategies, and is evaluated using proper scoring metrics, outperforming random sampling and cluster sampling approaches.
2023 IEEE BELGRADE POWERTECH
(2023)
Review
Computer Science, Artificial Intelligence
Thijs Peirelinck, Hussain Kazmi, Brida Mbuwir, Chris Hermans, Fred Spiessens, Johan Suykens, Geert Deconinck
Summary: Automated decision support tools utilizing machine learning and optimization algorithms have become crucial in addressing the increased electricity demand and variability associated with renewable energy sources. Recent advances have shown that transfer learning can significantly improve performance in this context.
Article
Engineering, Electrical & Electronic
Ehsan Rahmkhoda, Jawad Faiz, Moein Abedini
Summary: This paper investigates the failures and errors of loss-of-excitation (LOE) protection and proposes an intelligent method combined with a learning vector quantization (LVQ) neural network. The simulation results demonstrate that the proposed algorithm can detect LOE conditions quickly and prevent stresses applied to the machine and power system.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Andre M. Garcia, Leonardo H. Macedo, Ruben Romero
Summary: This article presents a specialized scatter search algorithm to solve the optimal transmission switching problem considering voltage control, and its efficiency is validated through tests.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Adel F. Alrasheedi, Khalid A. Alnowibet, Ahmad M. Alshamrani
Summary: This paper studies the bidding strategy problem of microgrid in a day-ahead electricity market. The DA electricity prices, demand, and uncertainties are modeled using different scenarios and robust optimization. A smart predict-and-optimize methodology is proposed, where a cost-oriented prediction model is used to replace accuracy-oriented methods. The trilevel mathematical programming approach is utilized to construct a cost-oriented model and solve the problem.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Depin Lv, Qiusheng Yu, Xiaoyong Wang, Yulian Bo, Yongjing Wei, Lei Liu, Pu Zhang, Yan Zhang, Wensheng Zhang
Summary: This study proposes a method based on the MHA-CNN framework to monitor the operation status of the smart grid and power communication network coupling. The method utilizes HHT to extract features from operational data and employs MHA to extract key feature information from fault data, achieving faster fault detection speed and higher detection probability.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Tao Zhang, Zhiguo Hao, Hongyue Ma, Songhao Yang, Chongtao Li
Summary: This paper introduces a dominant mode identification (DMI) method to assess the stability of grey-box grid-tied converters. The method analyzes the critical eigenvalues and can be applied to different grid connection scenarios.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Aamir Nawaz, Hongtao Wang, Huiting Yang, Hammad Armghan, Jiechao Gao
Summary: This paper presents a risk constrained probabilistic problem framework for addressing the challenges of stochastic dispatching due to contingencies and the integration of renewable sources. By utilizing probabilistic analytical target cascading, polynomial chaos expansion, and benders decomposition algorithm, the proposed method reduces computational burden and improves efficiency.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Thongchart Kerdphol, Issarachai Ngamroo, Tossaporn Surinkaew
Summary: This paper introduces an enhanced robust H-infinity technique to improve the regulation performance and cyber resiliency of secondary frequency control in a microgrid. The study shows that cyber-attacks affect different elements of secondary frequency control in the microgrid depending on the type of attack used.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Sambeet Mishra, Praveen Prakash Singh, Ivar Kiitam, Muhammad Shafiq, Ivo Palu, Chiara Bordin
Summary: This study aims to analyze partial discharge events in high-voltage cables and extract patterns and trends. The results indicate a positive correlation between the number of PD events and the increase in voltage levels, with negative PD peaks being more frequent at lower stress levels.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Cleberton Reiz, Jonatas B. Leite, Clara S. Gouveia, Mohammad Sadegh Javadi
Summary: This research proposes a bilevel method to simultaneously solve the allocation and coordination problems in microgrid protection, considering local devices. The results show that solving these problems simultaneously can save costs and select the optimal solution.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Rasha Elazab, M. Ser-Alkhatm, Maged A. Abu Adma, K. M. Abdel-Latif
Summary: This paper proposes a two-layer methodology for the allocation and sizing of SVCs in radial distribution systems. The Power Loss Index (PLI) is used to choose the best candidate buses for allocating SVCs, and a two-stage stochastic programming algorithm is proposed to solve the sizing problem. The advantages of combining SVCs and the PV inverter's reactive power provision ability are also analyzed.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Samreen Javaid, Dongyu Li, Abhisek Ukil
Summary: This paper introduces a new algorithm based on high pass filter for traveling wave fault location, which can differentiate the peak value of the traveling wave more easily by removing oscillation in the signal. The algorithm is tested on a DC simulation model and executed in MATLAB environment, achieving high accuracy in fault location results. Compared with traditional methods, this algorithm shows higher accuracy and faster response.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Enzhe Wang, Xiangrong Chen, Hanshan Zhu, Kai Yin, Muhammad Awais, Ashish Paramane
Summary: This paper aims to reduce the power loss of tunnel cable systems by using different grounding and loss reduction methods. The case study of the submarine tunnel power cable project in Ningbo-Zhoushan, China is considered. The nodal admittance matrices of the cable are established using the telegraph equation, and the cascading algorithm is used to analyze the induced voltage and circulating current in the cable sheath, taking into account the presence of grounding node. The study finds that coordinating the amplitude and phase angle of series impedance can reduce the power loss of the cable system.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Ziwei Ma, Jasronita Jasni, Mohd Zainal Abidin Ab Kadir, Norhafiz Azis
Summary: A numerical model was developed to analyze lightning attachment on UHV-AC double-circuit transmission lines. Simulation results showed that upward leaders simultaneously occurred on the ground wire and the upper-phase conductor when the phase conductors were energized. A new expression for calculating the striking distance on the ground wire was proposed based on the simulation results. Additionally, high voltage, phase angle, and terrain slope were found to increase the surface electric field of the upper-phase conductor.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Kunal Kumar, Susmita Kar
Summary: This paper presents an efficient protection scheme utilizing the rate of change parameter to detect both shunt faults and high impedance faults. By comparing with the existing scheme and testing on the IEEE 13 Bus test system, the effectiveness of the proposed scheme is demonstrated.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
Article
Engineering, Electrical & Electronic
Cheng Che, Bangbang He, Zhixuan Gao, Qiwei Lu, Yang Chen, Zhifeng Wang, Jinghan Guo, Yanwen Wang
Summary: A low-cost hybrid railway power conditioner (HRPC) technology scheme is proposed in this paper to effectively compensate negative sequence current (NSC) of the V/v traction system and reduce the capacity and cost of the railway power conditioner (RPC). The proposed HRPC combines an RPC with a modular multilevel converter structure (MMC-RPC) and a railway static var compensator (RSVC) and considers regenerative braking energy in the compensation process.
ELECTRIC POWER SYSTEMS RESEARCH
(2024)
