Search

• A hydrogen-based MEMG in the electricity market is exploited. • A multiple time-scale energy management solution for MEMG is developed. • Electricity and hydrogen trading among subsystems are considered. • Daily operational cost of the hydrogen-based MEMG is minimized. With the technological advances in energy conversion and utilization of multiple forms of energy sources, hydrogen energy has attracted increasing attention. The hydrogen energy can be used to directly supply the hydrogen demand and generate both electricity and heat through fuel cell-based combined heat and power (FC-CHP) unit. This paper proposed a multiple time-scale energy management solution for a hydrogen-based multi-energy microgrid (MEMG) to supply electricity, hydrogen and heating loads aiming to minimize the MEMG operational cost with consideration of renewable energy generation and demand uncertainties. The proposed solution consists of day-ahead energy scheduling and model predictive control (MPC) based real-time energy dispatch in the presence of the electricity market. In the hydrogen-based MEMG, the electricity and hydrogen can be dispatched and utilized across multiple interconnected subsystems to improve the overall system energy utilization efficiency. The proposed solution is extensively assessed through simulation experiments compared with a benchmark solution. The numerical results confirm that the proposed solution outperforms the benchmark solution with the mean daily actual operational costs reduced by 37.08%. [ABSTRACT FROM AUTHOR]

• A method for scenario analysis that takes into account the uncertainty of renewable energy was developed. • Two demand side response approaches were utilized to evaluate the response impacts of two different types of loads. • Carbon trading market was considered in the suggested microgrid model, and carbon emissions could be decreased by a maximum of 10.13 percent compared to case 1 when the market was not considered. • A multi-stage optimization framework for microgrid energy management was proposed. The benefits of more comprehensive energy use and promotion of renewable energy (RE) consumption enable large-scale microgrid deployment. However, the uncertainty of renewable energy sources and the diversity of load types pose a threat to the microgrid's stability. Recently, energy scheduling optimization for microgrids (MGs) has been primarily based on ideal models, but incorporating as many real-world characteristics as feasible can improve the reliability of the optimization outcome. This paper provides a multi-stage methodology for solving the energy management optimization (EMO) problem of MG under uncertainty considering carbon trading market and demand side response (DSR). To begin, scenario analysis method was used to address the uncertainty associated with RE in MG, and four typical scenarios of renewable energy were generated. Then, the flexible configuration and operational constraints of each power source in MG are dealt with under the premise of considering the carbon trading market. The third stage involved merging the characteristics of various load types and analyzing the response impacts of different percentage residential and industrial loads, respectively, using the price-based and load-transfer-based DSR approaches. Finally, quantum particle swarm optimization (QPSO) algorithm was used to obtain the optimal solution. The acquired results demonstrate the efficacy of the proposed multi-stage energy optimization framework, and support the following two conclusions: 1) The carbon trading market policy contributes to the reduction of carbon emissions and fossil fuel consumption. 2) A high load participation rate in DSR can increase MG operation economics by up to 27.48% compared to not considering DSR. [ABSTRACT FROM AUTHOR]

This paper proposes an energy management strategy for a residential microgrid comprising photovoltaic (PV) panels and a small wind turbine. The microgrid is connected to the main grid, allowing for a controlled power exchange through a battery system and its control strategy. As input data, the proposed control strategy uses the battery state of charge (SOC), the power at each microgrid node as well as the load and renewable generation forecasts. By using forecasted data and correcting any forecasting errors according to the SOC of the battery, the strategy manages to make a better use of the battery resulting in a better grid power profile. The simulation of the system using a one-year data period shows that the proposed energy management strategy results in a better grid power profile for a given storage system when compared with other state-of-the-art strategies. Finally, the proposed strategy was experimentally validated in the microgrid built in the Renewable Energy Laboratory at the UPNa. [ABSTRACT FROM AUTHOR]

To reduce global greenhouse gas emissions, the world must find intelligent solutions to maximise the utilisation of carbon-free renewable energy sources. In this paper, multi-agent reinforcement learning is used to control a microgrid in a mixed cooperative and competitive setting. The agents observe fluctuating energy demand, dynamic wholesale energy prices, and intermittent renewable energy sources to control a hybrid energy storage system to maximise the utilisation of the renewables to reduce the energy costs of the grid. In addition, an aggregator agent trades with external microgrids competing against one another and the aggregator to reduce their own energy bills. For this, the algorithm deep deterministic policy gradients (DDPG) and multi-agent DDPG (MADDPG) are used to compare the use of a single global controller versus multiple distributed agents, along with the single and multi-agent variants of distributional DDPG (D3PG) and twin delayed DDPG (TD3). The research found it is significantly more profitable for the primary microgrid to sell energy on its own terms rather than selling back to the utility grid, and is also beneficial for the external microgrids as they also reduce their own energy bills. The methods that produced the greatest profits were the multi-agent approaches where each agent has its own reward function based on the principle of marginal contribution from game theory. The multi-agent approaches were better able to evaluate their performance controlling individual components of the environment which allowed them to develop their own unique policies for the different types of energy storage system. • Novel use of MADDPG for mixed cooperative and competitive microgrid environment. • Compare use of a central controller or multiple distributed agents in microgrids. • Hybrid energy storage system for long, medium, and short-term storage. • Case studies for renewable utilisation and trading with neighbouring microgrids. • Showcase the benefit of using rewards based on marginal contribution in a MAS. [ABSTRACT FROM AUTHOR]

• Cloud computing technologies applied to microgrid management. • Supervised machine learning strategies associated with the energy management system. • Used architecture by the energy management system to control a cluster of microgrids. • Considered elements in developing a scalable and autonomous architecture. The way to organize the generation, storage, and management of renewable energy and energy consumption features has taken relevance in recent years due to demands that define the social welfare of this century. Like demand increases, other factors require grid infrastructure improvement, updates, and opening to other technologies that assuage the final customer needs. Precisely, the interest in renewable energy sources, the constant evolution of energy storage technologies, the continuous research involving microgrid management systems, and the evolution of cloud computing technologies and machine learning strategies motivate the development of this article. Tasks associated with a microgrid cluster like the integration of a considerable number of heterogeneous devices, real-time support, information processing, massive storage capabilities, security considerations, and advanced optimization techniques usage could take place in an autonomous and scalable energy management system architecture under a machine learning perspective running in real-time and using Cloud resources. This paper focuses on identifying the elements considered by different authors to define a cloud-based architecture and ensure the appropriately supervised learning functionality under a microgrids cluster environment. Namely, it was necessary to revise and run microgrid simulations, real-time simulation platforms usage, connection to a virtual server for microgrid control and set the energy management system using cloud computing and machine learning. Based on the review and considering the scenarios mentioned, this article presents a scalable and autonomous cloud-based architecture that allows power generation forecast, energy consumption prediction, a real-time energy management system using machine learning techniques. [ABSTRACT FROM AUTHOR]