Uma abordagem orientada a eventos para o planejamento de recursos em sistemas de geração distribuída de energia

Abstract

Renewable power generation systems stand as astraightforward alternative to complement the energy grid of the future. Physically, generation components have distributed nature and they work individually to supply centralised microgrids, which interface them with the main power grid and consumers. As microgrid components are non-cooperative entities, engineers are limited to plan resources consumption, fulfil unit commitments, stratify time, cost, environmental impact, quality of energy, etc. In this paper, we exploit the event-driven behaviour of microgrids to provide ways for them to recognise, communicate, and cooperate with each other towards common goals. We first design microgrids and generators as self-controlled cyber-physical agents that communicate with each other and with the main power grid. Then, we develop a Petri net simulation model that estimates customisation policies to be applied backover the microgrid via event-based control. Results show that in some cases it is possible to save effort (and resources) of power generation by replacing reactiveness for a more balanced, context-aware, energy chain. A real photovoltaic plant is modelled by our approach with 86% accuracy.