Electric Distribution Network With Multi-Microgrids Management Using Surrogate-Assisted Deep Reinforcement Learning Optimization

Kaewdornhan, Niphon; Srithapon, Chitchai; Chatthaworn, Rongrit

doi:10.1109/access.2022.3229127

Cited by 10 publications

(3 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, authors in [38] adopt multi-agent reinforcement learning to minimize peak EV charging demand at EVCS. In [39], authors use DRL to decide the real-time price to dispatch multi microgrids in a power distribution network. Authors in [40] adopt Actor3Critic A3C-LSTM DRL to determine the retail price to maximize social profits in EV dispatching communities.…”

Section: A Cooperation Of the Power Network And The Transportation Ne...mentioning

confidence: 99%

Price-Guided Cooperation of Combined Offshore Wind and Hydrogen Plant, Hydrogen Pipeline Network, Power Network, and Transportation Network

Li,

Jian

2023

IEEE Access

View full text Add to dashboard Cite

Nowadays, the combined offshore wind farm and hydrogen production plant (COWHP) has been installed all around the world and will play an important role in reducing the carbon dioxide emissions. However, the cooperation of the COWHP, the hydrogen pipeline transportation, the power network transmission, and the hydrogen fuel cell electric vehicles (HFCEVs) scheduling is still an essential problem. This problem falls into the category of mixed-integer nonlinear optimization, which involves a significant number of decision variables and constraints. Existing methods have often struggled to effectively solve this problem due to its inherent complexity and non-linear nature. In this paper, a priceguided method is proposed to cooperate with this complex system. First, the COWHP model, the hydrogen pipeline network scheduling model, the power network flow model, the HFCEVs refuelling microgrid model, and the HFCEV traffic flow model in real-world transportation networks are presented. Second, the price-based cooperation model is proposed. Third, price strategies are deployed to cooperate with the complex system. The decision making trial and evaluation laboratory-the technique for order preference by similarity to ideal solution (Dematel-TOPSIS) method is deployed to evaluate the real-time performance of different strategies. The simulation results reveal that in the small vehicle flow transportation network cases, deep deterministic policy gradient (DDPG) has the best real-time evaluation performance; whereas for the large vehicle flow transportation network cases, long short term memory (LSTM) has the best real-time performance. From the posterior evaluation view, LSTM has the best performance for all cases to reduce total operation cost, and also reduce the total waiting time.INDEX TERMS offshore wind farm, hydrogen production, hydrogen pipeline transportation, microgrid, hydrogen fuel cell electric vehicles, transportation network.

show abstract

Section: A Cooperation Of the Power Network And The Transportation Ne...mentioning

confidence: 99%

Price-Guided Cooperation of Combined Offshore Wind and Hydrogen Plant, Hydrogen Pipeline Network, Power Network, and Transportation Network

Li,

Jian

2023

IEEE Access

View full text Add to dashboard Cite

show abstract

“…This power use should be scheduled so that the user does not wait long and there is no surge in power use. Scheduling from battery charging can be charging at a time that not many users use [3] [4] [5], combined with renewable energy [6] [7] [8], or trying to increase profits [9] [10]. The scheduling method mentioned above requires an accurate prediction of battery availability to be implemented correctly.…”

Section: Introductionmentioning

confidence: 99%

LSTM-based forecasting on electric vehicles battery swapping demand: Addressing infrastructure challenge in Indonesia

Romdlony,

Khayr,

Muharam

et al. 2023

J. Mechatron. Electr. Power Veh. Technol.

View full text Add to dashboard Cite

This article aims to design a model for forecasting the number of vehicles arriving at the battery swap station (BSS). In our case, we study the relevance of the proposed approach given the rapid increase in electric vehicle users in Indonesia. Due to the vehicle electrification program from the government of Indonesia and the lack of supporting infrastructure, forecasting battery swap demands is very important for charging schedules. Forecasting the number of vehicles is done using machine learning with the long short-term memory (LSTM) method. The method is used to predict sequential data because of its ability to review previous data in addition to the current input. The result of the forecasting using the LSTM method yields a prediction score using the root-mean-square error (RMSE) of 2.3079 x 10-6 . The forecasted data can be combined with the battery charging model to acquire predicted hourly battery availability that can be processed further for optimization and scheduling.

show abstract

“…Uncertainties were modeled intelligently using historical data. Reference [46] presented a two-level control of MMGs using deep neural networks. The computational burden was reduced by estimating the parameters of MGs, instead of calculating the probabilistic power flow.…”

Section: Introductionmentioning

confidence: 99%

A Novel Multi Level Dynamic Decomposition Based Coordinated Control of Electric Vehicles in Multimicrogrids

Aslam,

Kashif,

Gulzar

et al. 2023

Sustainability

View full text Add to dashboard Cite

This paper presents a novel tetra-level dynamic decomposition-based control approach for coordinated operation of electric vehicles in multimicrogrids, which is comprehensive, generic, modular, and secure in nature, to maximize the utilization of renewable energy sources, while meeting the load demands with the resources available. There are a number of microgrids that are connected to the grid. Each microgrid consists of a number of renewable energy sources, energy storage systems, non-renewable energy sources, electric vehicles, and loads. Each distributed energy source or load is controlled by a microsource controller. All microsource controllers with a similar nature are controlled by a unit controller, and all the unit controllers in a microgrid are controlled by a microgrid controller. There is a single multimicrogrid controller at the top. The proposed control scheme was verified through simulation-based case studies.

show abstract

Electric Distribution Network With Multi-Microgrids Management Using Surrogate-Assisted Deep Reinforcement Learning Optimization

Cited by 10 publications

References 56 publications

Price-Guided Cooperation of Combined Offshore Wind and Hydrogen Plant, Hydrogen Pipeline Network, Power Network, and Transportation Network

Price-Guided Cooperation of Combined Offshore Wind and Hydrogen Plant, Hydrogen Pipeline Network, Power Network, and Transportation Network

LSTM-based forecasting on electric vehicles battery swapping demand: Addressing infrastructure challenge in Indonesia

A Novel Multi Level Dynamic Decomposition Based Coordinated Control of Electric Vehicles in Multimicrogrids

Contact Info

Product

Resources

About