 Deep learning has become increasingly popular in recent years due to its ability to process large amounts of data quickly and accurately. In this study, five different deep learning models were evaluated for their ability to accurately predict short-term PV power generation in a real-world floating PV power plant. The CNN by GRU model was found to be the most accurate for one-day predictions, while the by LSTM model was the most accurate for one-week predictions. These findings suggest that deep learning techniques can provide more accurate short-term PV power generation forecasts than traditional predictive methods like linear regression. This article was authored by Nantawa Kortsriwan, Promfak Brunraksa, Tarapung Brunraksa, and others.