Twin Delayed Deep Deterministic Policy Gradient (TD3) Based Virtual Inertia Control for Inverter-Interfacing DGs in Microgrids

Egbomwan, Osarodion E.; Liu, Shichao; Chaoui, Hicham

doi:10.1109/jsyst.2022.3222262

Cited by 10 publications

(7 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This additional controller, implementing VI, is integrated into the microgrid system, utilising model predictive control (MPC) with a response mechanism for VI. Using the TD3 scheme, an estimation method for inertia level is demonstrated in [105].…”

Section: Intelligent Strategiesmentioning

confidence: 99%

Virtual Inertia Control for Power Electronics-Integrated Power Systems: Challenges and Prospects

Shobug,

Chowdhury,

Hossain

et al. 2024

Energies

View full text Add to dashboard Cite

In modern power systems, conventional energy production units are being replaced by clean and environmentally friendly renewable energy resources (RESs). Integrating RESs into power systems presents numerous challenges, notably the need for enhanced grid stability and reliability. RES-dominated power systems fail to meet sufficient demand due to insufficient inertia responses. To address this issue, various virtual inertia emulation techniques are proposed to bolster power system stability amidst the increased integration of renewable energy sources into the grid. This review article explores state-of-the-art virtual inertia support strategies tailored to accommodate the increased penetration of RESs. Beginning with an overview of this study, it explores the existing virtual inertia techniques and investigates the various methodologies, including control algorithms, parameters, configurations, key contributions, sources, controllers, and simulation platforms. The promising virtual inertia control strategies are categorised based on the techniques used in their control algorithms and their applications. Furthermore, this review explains evolving research trends and identifies promising avenues for future investigations. Emphasis is placed on addressing key challenges such as dynamic response characteristics, scalability, and interoperability with conventional grid assets. The initial database search reveals 1529 publications. Finally, 106 articles were selected for this study, adding 6 articles manually for the review analysis. By synthesising current knowledge and outlining prospective research directions, this review aims to facilitate the current state of research paths concerning virtual inertia control techniques, along with the categorisation and analysis of these approaches, and showcases a comprehensive understanding of the research domain, which is essential for the sustainable integration of renewable energy into modern power systems via power electronic interface.

show abstract

Section: Intelligent Strategiesmentioning

confidence: 99%

Virtual Inertia Control for Power Electronics-Integrated Power Systems: Challenges and Prospects

Shobug,

Chowdhury,

Hossain

et al. 2024

Energies

View full text Add to dashboard Cite

show abstract

“…Additionally, TD3 applies random noise to the target policy during training, which encourages exploration and can lead to better generalization. 88 In recent years, TD3 has been applied to a variety of real-world problems with promising results. For example, in reference [89], the authors propose a TD3-based controller for the PA in the Internet of Robotic Things (IoRT) network.…”

Section: Twin Delayed Deep Deterministic Policy Gradient (Td3)mentioning

confidence: 99%

“…One key feature of TD3 is the use of two critics instead of one, which helps to better estimate the Q ‐values. Additionally, TD3 applies random noise to the target policy during training, which encourages exploration and can lead to better generalization 88 . In recent years, TD3 has been applied to a variety of real‐world problems with promising results.…”

Section: Literature Reviewmentioning

confidence: 99%

Resource allocation in 5G cloud‐RAN using deep reinforcement learning algorithms: A review

Khani,

Jamali,

Sohrabi

et al. 2023

Trans Emerging Tel Tech

View full text Add to dashboard Cite

This paper reviews recent research on resource allocation in 5G cloud‐based radio access networks (C‐RAN) using deep reinforcement learning (DRL) algorithms. It explores the potential of DRL for learning complex decision‐making policies without human intervention. The paper first introduces the C‐RAN architecture and resource allocation concepts, followed by an overview of DRL algorithms applied to C‐RAN. It discusses the challenges and potential solutions in applying DRL to C‐RAN resource allocation, including scalability, convergence, and fairness. The review concludes by highlighting open research directions for future investigation. By providing insights into the state‐of‐the‐art techniques for resource allocation in 5G C‐RAN using DRL, this paper emphasizes their potential impact on advancing 5G network technology.

show abstract

“…The probability of picking an action a t at a given state s t is represented through the policy π a t js t ð Þ. The goal of the agent is to achieve the optimal policy π Ã ,which will maximize the total long-term anticipated reward R. The assessment of benefit in reaching a particular state is determined through the state value function V π s t ð Þ 35 and is expressed as…”

Section: Working With Td3pg Control Agentsmentioning

confidence: 99%

Fault‐resilient control of parallel PV inverters using multi‐agent twin‐delayed deep deterministic policy gradient approach

Malik,

Haque,

Kurukuru

et al. 2023

Circuit Theory & Apps

View full text Add to dashboard Cite

SummaryGrid‐tied photovoltaic system (GTPS) are widely favored due to their inherent benefits. The interface of parallel power electronic converters in these systems is rapidly increasing. The inverter in GTPS has a very important role in power conversion, transfer, and control. However, their challenging working conditions contribute to susceptibility to power switching device failures. Traditional fault diagnosis and tolerant approaches exhibit limitations in control efficiency, model dependency, and slow recovery from fault. This study proposes a multi‐agent twin delayed deep deterministic policy gradient (MATD3PG) configuration for intelligent parallel inverter control, fault diagnosis, and fault‐tolerant operation in GTPS structure. By leveraging the multi‐agent reinforcement learning (RL) framework, an optimal control of the parallel inverter can be achieved, encompassing fault‐tolerant operation using MATLAB Simulink/PLECS. The major advantage the given technique offers is that it carries out optimum fault‐tolerant operation without causing the system derating. Experimental findings demonstrate that the proposed fault‐tolerant model based on RL traditional methods is able to ensure continuous supply to the connected loads even during fault events. Further, the transition time from fault occurrence to recovery is found to be 6 ms, which is quite less compared with the fault‐tolerant techniques presented in literature. Through real‐time fault diagnosis‐based results, the proposed approach ensures precise tracking of reference currents, quicker response times, uninterrupted supply, and smooth transition to the post‐fault operation mode.

show abstract

Twin Delayed Deep Deterministic Policy Gradient (TD3) Based Virtual Inertia Control for Inverter-Interfacing DGs in Microgrids

Cited by 10 publications

References 43 publications

Virtual Inertia Control for Power Electronics-Integrated Power Systems: Challenges and Prospects

Virtual Inertia Control for Power Electronics-Integrated Power Systems: Challenges and Prospects

Resource allocation in 5G cloud‐RAN using deep reinforcement learning algorithms: A review

Fault‐resilient control of parallel PV inverters using multi‐agent twin‐delayed deep deterministic policy gradient approach

Contact Info

Product

Resources

About