Design and planning of flexible mobile Micro-Grids using Deep Reinforcement Learning

Ongoing risks from climate change have significantly impacted the livelihood of global nomadic communities and are likely to lead increased migratory movements in coming years. As a result, mobility considerations becoming increasingly important energy systems planning, particularly achieve access developing countries. Advanced “Plug Play” control strategies been recently developed with such decentralized framework mind, allowing easier interconnection communities, both each other main grid. Considering above, design planning strategy mobile multi-energy supply system for community is investigated this work. Motivated by scale dimensionality associated uncertainties, impacting all major decision variables over 30-year horizon, Deep Reinforcement Learning (DRL) Flexibility Analysis implemented problem. DRL based solutions benchmarked against several rigid baseline options compare expected performance under uncertainty. The results on case study ger Mongolia suggest that can be technically economically feasible, when considering flexibility, although degree spatial dispersion among households an limiting factor. Additionally, policies development dynamic evolution adaptability strategies, which used targeted very wide range potential scenarios. Key economic, sustainability resilience indicators as Cost, Equivalent Emissions Total Unmet Load measured, suggesting improvements compared available baselines up 25%, 67% 76%, respectively. Finally, decomposition values flexibility plug play operation presented using variation real theory, implications policymakers focused enabling their access.