.png){kind=icon}
.png){kind=icon}
Return
DeepComp: Deep reinforcement learning based renewable energy error compensable forecasting
.png){kind=icon}
PRE
AI
DOI：10.1016/j.apenergy.2021.116970.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Recently, renewable energy is rapidly integrated into the power grid to prevent climate change, and accurate forecasting of renewable generation becomes critical for reliable power system operation. However, existing forecasting algorithms only focused on reducing forecasting errors without considering error compensability by using a large-scale battery. In this paper, we propose a novel strategy called error compensable forecasting. We switch the objective of forecasting from reducing errors to making errors compensable by leveraging a battery, which in turn reduces the dispatched error, the difference between forecasted value and dispatched value. The challenging part of the proposed objective lies in that the stored energy at current time is affected by the previous forecasting result. In this regard, we propose a deep reinforcement learning based error compensable forecasting framework, called DeepComp, having forecasting in the loop of control. This makes an action as a continuous forecasted value, which requires a continuous action space. We leverage proximal policy optimization, which is simple to implement with outstanding performance for continuous control. Extensive experiments with solar and wind power generations show that DeepComp outperforms the conventional forecasting methods by up to 90% and achieves accurate forecasting, e.g., 0.58-1.22% of the mean absolute percentage error.
Keywords:
Deep reinforcement learning
Proximal policy optimization
Renewable energy forecasting
Battery control
Error compensation
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.