Artificial intelligence-based hierarchical management for optimized performance and energy trading in networked microgrids

With the increasing demand for electrical energy due to population growth and technological advancements, coupled with concerns over fossil fuel prices and environmental impacts, the integration of renewable energy sources like photovoltaic and wind turbines has become imperative. Microgrid offers a solution by efficiently meeting load demands through a combination of renewable energy sources and conventional resources like diesel generators and battery storage systems. For enhanced utility, adjacent microgrid s are interconnected to form Networked Microgrid systems. The optimal integration of networked microgrid systems necessitates addressing three fundamental aspects: optimizing microgrid size, effectively managing diverse microgrid resources, and facilitating energy trading within local energy markets. This thesis presents a two-level hierarchical strategy based on artificial intelligence methods to address these problems. A hybrid artificial intelligence method is developed by integrating grey wolf optimizer and particle swarm optimization technique called hybrid grey wolf- particle swarm optimization to determine optimal configurations and sizes for microgrid. Multi-objective optimization, considering levelized cost of energy, power losses, and gas emissions, is employed and reliability indicators including losses of power supplied probability, cost benefit index, and participation of renewable index are monitored. The results showed that the configuration that include photovoltaic-wind turbines-diesel generators-battery storage systems delivered highly favorable outcomes. Specifically, it achieved an impressive levelized cost of energy of 0.127 $/kWh, correspondingly, the losses of power supplied probability index was (0.0271). A comparative analysis was conducted with results obtained from the firefly algorithm, particle swarm optimization, and grey wolf optimizer techniques individually, highlighting the superiority of the hybrid grey wolf- particle swarm optimization over the others. Furthermore, this thesis proposes a distributable resource management strategy to reduce gas emissions and minimize levelized cost of energy. Additionally, a demand-side management strategy is introduced to shift non-sensitive loads from periods of energy shortage to surplus. These dual strategies succeeded in diminishing the levelized cost of energy and gas emissions of Microgrids 1, 2, 3, and 4 to (0.125, 0.127, 0.129, 0.128) $/kWh and (5.77, 5.52, 19.98, 14.79) tone/year respectively. Finally, a revised approach for energy trading in the local energy market is presented, encompassing various mechanisms including peer-to-peer, peer-to-grid, and distributed methods. Multi-round energy auctions, considering transmission line capacities and costs, determine energy prices and quantities traded. peer-to-peer yields the highest annual benefit, while distributed trading reduces losses of power supplied probability. Overall, the proposed strategies prove effective in enhancing Networked Microgrid systems performance, showcasing significant improvements in economic, technical, and environmental aspects. These contributions collectively demonstrate significant improvements in the economic, technical, and environmental performance of Networked Microgrid systems, making them more sustainable and efficient for future energy demands.

Text FK 2024 48 - Declaration Form.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (930kB) Text FK 2024 48 - Full Text.pdf Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (3MB) Text FK 2024 48.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (3MB) Official URL or Download Paper: http://ethesis.upm.edu.my/id/eprint/19013

Actions (login required)

View Item