Self-learning Prosumer in Competitive Local Energy Market

In this paper, we study a prosumer’s energy trading behavior in a proposed local energy market (LEM). The prosumer will try to make the best use of its available energy resources, such as its PV panel and battery, with the self-learning capability to make a smart operation strategy. By adapting a deep reinforcement learning (DRL) technique and a rainflow battery aging mechanism, it can be taken as a contribution that the proposed intelligent and economical decision-making framework will facilitate more customers while encouraging active participation in the localized energy ecosystem. Meanwhile, more energy business models can be strategically redesigned on top of it.