Q-learning algorithm based multi-agent coordinated control method for microgrids

Yuanyuan, Xi; Chang, Liuchen; Mao, Meiqin; Jin, Pengfei; Hatziargyriou, Nikos D.; Xu, Hao

doi:10.1109/icpe.2015.7167977

Cited by 6 publications

(4 citation statements)

References 13 publications

(12 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Noting that the action that gives the most reward is the greedy action, , strategies for solving the dilemma of exploration and exploitation have been developed, apart from random selection, namely; -greedy [31], the pursuit algorithm [64], the SoftMax function [30], random [65] and deterministic [45].…”

Section: B Exploration Exploitation Conundrummentioning

confidence: 99%

See 1 more Smart Citation

Reinforcement Learning Techniques for Optimal Power Control in Grid-Connected Microgrids: A Comprehensive Review

Arwa

Folly

2020

IEEE Access

View full text Add to dashboard Cite

The utility grids are undergoing several upgrades. Distributed generators that are supplied by intermittent renewable energy sources (RES) are being connected to the grids. As RES generations get cheaper, more customers are opting for peer-to-peer energy interchanges through the smart metering infrastructure. Consequently, power management in grid-tied RES-based microgrids has become a challenging task. This paper reviews the applications of reinforcement learning (RL) algorithms in managing power in grid-tide microgrids. Unlike other optimization methods such as numerical and soft computing techniques, RL does not require an accurate model of the optimization environment in order to arrive at an optimal solution. In this paper, various challenges associated with the control of power in grid-tied microgrids are described. The application of RL techniques in addressing those challenges is reviewed critically. This review identifies the need to improve and scale multi-agent RL methods to enable seamless distributed power dispatch among interconnected microgrids. Finally, the paper gives directions for future research, e.g., the hybridization of intrinsic and extrinsic reward schemes, the use of transfer learning to improve the learning outcomes of RL in complex power systems environments and the deployment of priority-based experience replay in post-disaster microgrid power flow control. INDEX TERMS Electric vehicle charging station, energy management, Markov Decision Process, microgrid, reinforcement learning.

show abstract

Section: B Exploration Exploitation Conundrummentioning

confidence: 99%

“…The algorithm converged to a policy that reduced energy interchange with the grid by 14%. Other applications of the use of Q-learning in energy scheduling in microgrids may be found in [65], [71] and [72].…”

Section: Q-learningmentioning

confidence: 99%

Reinforcement Learning Techniques for Optimal Power Control in Grid-Connected Microgrids: A Comprehensive Review

Arwa

Folly

2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…In the second scenarios, the BG2 is involved at 0.8 s and the first-level control will be activated at t = 1 s. The reactive power demand of DG 1 0 is decreased by 15 kVar at t = 1 s. It can be observed that the fluctuation of voltage can be eliminated because of the consensus protocol (12)- (14). The first-level control and the second-level control will not impact each other and the reactive power is decreased by 15 kVar at t = 2 s, t = 3 s to verify the validation.…”

Section: Performance Validationmentioning

confidence: 99%

“…The method based on Qlearning and graph theory is aimed to degrade the power grid pressure and increase the power grid stability. Q-learning is utilised for setting coordination control objectives and accelerating convergence characteristic [14]. On the other hand, Q-learning can optimise the power scheduling and increase the economy [15,16].…”

Section: Introductionmentioning

confidence: 99%

Distributed cooperative control method based on network topology optimisation in microgrid cluster

Dou

Yue

et al. 2020

IET Renewable Power Generation

View full text Add to dashboard Cite

When several microgrids are interconnected into microgrid cluster (MGC), the network topology of MGC system would be more complex. Traditional hierarchical control based on droop control tends to make full use of all networks so that the costs of energy exchange could be high. Considering the voltage stability and power sharing of MGC system, a distributed cooperative control method based on network topology optimisation for MGC is proposed to control the power flow of multimicrogrid system. Specially, preliminary Q-leaning algorithm is utilised with establishment of new adjacency matrices for the distributed cooperative control to decrease the costs. Moreover, multi-level controls based on consensus protocol are proposed in the distributed cooperative MGC system to eliminate the necessity of a central controller. The adaptive regulation based on the state of charge among the battery groups will be taken as the first-level control. The second-level control including the current controller, voltage controller and power controller is implemented among the DGs and a novel third-level control based on consensus is applied in eliminating the voltage deviation caused by the first-level control and the second-level control. The effectiveness of the proposed method is verified through the simulation by simulink.

show abstract

Power Flow Management in Multi-Source Electric Vehicle Charging Station

Erick

Folly

2020

IFAC-PapersOnLine

View full text Add to dashboard Cite

Q-learning algorithm based multi-agent coordinated control method for microgrids

Cited by 6 publications

References 13 publications

Reinforcement Learning Techniques for Optimal Power Control in Grid-Connected Microgrids: A Comprehensive Review

Reinforcement Learning Techniques for Optimal Power Control in Grid-Connected Microgrids: A Comprehensive Review

Distributed cooperative control method based on network topology optimisation in microgrid cluster

Power Flow Management in Multi-Source Electric Vehicle Charging Station

Contact Info

Product

Resources

About