Multiagent-Based Hybrid Energy Management System for Microgrids

Mao, Meiqin; Jin, Pengfei; Hatziargyriou, Nikos D.; Chang, Liuchen

doi:10.1109/tste.2014.2313882

Cited by 112 publications

(79 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…MAS techniques are also increasingly used in energy Demand Side Management (DSM) solutions to reduce peak power demand and thus minimize CO 2 emissions Logenthiran et al, 2011), and better integrate renewable energy sources (Gomes et al, 2014;Mao et al, 2014). There are also important financial implications to DSM, as customers (and generators) can reduce their overall costs.…”

Section: Multi-agent Systems In Smart Gridsmentioning

confidence: 99%

Prediction-Based Multi-Agent Reinforcement Learning in Inherently Non-Stationary Environments

Marinescu

Dusparić

Clarke

2017

ACM Trans. Auton. Adapt. Syst.

View full text Add to dashboard Cite

Multi-agent reinforcement learning (MARL) is a widely researched technique for decentralised control in complex large-scale autonomous systems. Such systems often operate in environments that are continuously evolving and where agents’ actions are non-deterministic, so called inherently non-stationary environments. When there are inconsistent results for agents acting on such an environment, learning and adapting is challenging. In this article, we propose P-MARL, an approach that integrates prediction and pattern change detection abilities into MARL and thus minimises the effect of non-stationarity in the environment. The environment is modelled as a time-series, with future estimates provided using prediction techniques. Learning is based on the predicted environment behaviour, with agents employing this knowledge to improve their performance in realtime. We illustrate P-MARL’s performance in a real-world smart grid scenario, where the environment is heavily influenced by non-stationary power demand patterns from residential consumers. We evaluate P-MARL in three different situations, where agents’ action decisions are independent, simultaneous, and sequential. Results show that all methods outperform traditional MARL, with sequential P-MARL achieving best results.

show abstract

Section: Multi-agent Systems In Smart Gridsmentioning

confidence: 99%

Prediction-Based Multi-Agent Reinforcement Learning in Inherently Non-Stationary Environments

Marinescu

Dusparić

Clarke

2017

ACM Trans. Auton. Adapt. Syst.

View full text Add to dashboard Cite

show abstract

“…Where (2) is the output of the j-th node in HL, wij are the connective weights from the input nodes to hidden nodes,…”

Section: (K)-p(k) and Its Derivative Dep(k)=ep(k)-ep(k-1) Whereas Itmentioning

confidence: 99%

“…Where (1) is the i-th input of FFNN, (1) ( ) is the i-th output of the IL, (2) ( ) is the output of the HL, (3) is the input of the output neuron, (3) ( ) is the output of FFNN, wij means the weight between the i-th node of the IL and j-th node of the HL, woj is the weight connecting the j-th node of the HL to the OL, and (2) , (3) are the TS-AFs for the HL and the OL respectively. The control output signal of the FFNN is multiplied by the measured DC-bus voltage to determine the active power reference P * (k).…”

Section: Ffnn Based DC Voltage Controllermentioning

confidence: 99%

“…Many works have focused on hybrid microgrids and have proposed a number of control schemes for different mode of operations [2]- [8]. A multiagent-based energy management system to optimizes the economic operation of a MG is presented in [2].…”

mentioning

confidence: 99%

“…A multiagent-based energy management system to optimizes the economic operation of a MG is presented in [2]. A reactive power sharing algorithm in hierarchical droop control is developed in [3].…”

mentioning

confidence: 99%

See 2 more Smart Citations

Adaptive Neural Network-Based Control of a Hybrid AC/DC Microgrid

Chettibi

Mellit

Sulligoi

et al. 2016

IEEE Trans. Smart Grid

View full text Add to dashboard Cite

-In this paper, the behavior of a grid-connected hybrid AC/DC Microgrid has been investigated. Different Renewable Energy Sources -photovoltaics modules and a wind turbine generator -have been considered together with a Solid Oxide Fuel Cell and a Battery Energy Storage System. The main contribute of this work is the design and the validation of an innovative online-trained artificial neural network based control system for a hybrid microgrid. Adaptive Neural Networks are used to track the Maximum Power Point of renewable energy generators and to control the power exchanged between the Front-End Converter and the electrical grid. Moreover, a fuzzy logic based Power Management System is proposed in order to minimize the energy purchased from the electrical grid. The operation of the hybrid microgrid has been tested in the Matlab/Simulink environment under different operating conditions. The obtained results demonstrate the effectiveness, the high robustness and the self-adaptation ability of the proposed control system. OWADAYS, the wide diffusion of distributed RES presents a new scenario for the regulation of distribution networks and the availability of new technologies for storage systems encourages their use in power systems [1]. In general, a hybrid AC/DC MG integrates different Distributed Generators (e.g. solar power sources, wind power generators, cogenerators, etc.), a energy storage system and a number of AC and DC loads. A FEC can interface the MG with the electric grid and can operate either in a grid-connected or islanded mode. The use of a PMS is crucial to optimize the power flow through the different components of the MG and the exchange of energy with the electric grid. Moreover, since the power produced by RESs depends on the climatic conditions, MPPT algorithms are needed in order to harvest the maximum available energy. The intermittent nature of RESs with the time-varying loads demand make the use of advanced control structures fundamental in order to make the operation of the MG reliable, economic, and secure under different operating conditions. The MG must also guarantee a high quality power supply to both local loads and electrical grid. Index TermsMany works have focused on hybrid microgrids and have proposed a number of control schemes for different mode of operations [2]- [8]. A multiagent-based energy management system to optimizes the economic operation of a MG is presented in [2]. A reactive power sharing algorithm in hierarchical droop control is developed in [3]. A novel coordinated voltage control scheme with islanding capability for a MG is proposed in [4]. adapt to uncertainties and they can be used also when the model of the system to be controlled is not available. Recently, the NNs with the learning capability are widely applied for the control of complex power systems. In [9], a back-propagation NN is applied for the real-time estimation of the wind speed. A novel discrete-time NN controller for the control of DC distribution system is designed in [10]. In [11], a...

show abstract