Intelligent management of the Operation of a Photovoltaic Plant based on Deep Reinforcement Learning
Deep Reinforcement Learning. Free Energy Market. Energy Storage System. Photovoltaic Plant.
In the Electric Power commercialization on the Free Contracting Environment (ACL), the producers that generates their energies from renewable sources (especially windy and solar), they will face uncertainties of how increase their revenues due uncertainties resources from environment, like the wind and solar radiation, in addition to uncertainties of electricity prices. This work proposes a Deep Reinforcement Learning (DRL) method to address this issue. According to this method, an agent, modeling by a Deep Learning network, must be trained to map the input states and control the actions of a Photovoltaic Plant (UFV) equipped with an Energy Storage System (SAE). By the agent, the influence of the uncertainties of solar generation and electricity prices on the revenue can be automatically involved and an expected optimal decision can be obtained. The work aims to overcome the challenges of uncertainties and optimal revenues for the Photovoltaic Plant.