Portal de Periódicos da CAPES

Predicting Renewable Energy Generation Using LSTM for Risk Assessment of Local Level Power Networks

2020; Korean Institute of Electrical Engineers; Volume: 69; Issue: 6 Linguagem: Inglês

10.5370/kiee.2020.69.6.783

2287-4364

Ho-Sung Ryu, Yong-Rae Lee, Mun-Kyeom Kim,

Integrated Energy Systems Optimization

Low uncertainty is essential when operating the power system in a stable state. Recently, the uncertainty in the power systems has increased due to the growth of renewable energy. This paper proposes a method to reduce the uncertainty of the power systems including renewable energy by using Long Short-term Memory (LSTM) algorithm. Through repeated simulation, the optimal LSTM model of each renewable unit is created. probabilistic scenario is created by monte-carlo simulation and k-means clustering algorithm, and then we assess risk for each scenario through a test system created with reference to the actual system. To validate the superiority of the proposed method, the risk assessment are conducted through local level test system. The results demonstrate that the optimal LSTM model reduces the risk index compared to other predicted models.