February 8, 2024Results in EngineeringOpen Access

Bi-LSTM, GRU and 1D-CNN models for short-term photovoltaic panel efficiency forecasting case amorphous silicon grid-connected PV system

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Empirical analysis demonstrates high accuracy in forecasting PV power using advanced deep learning models, indicating practical benefits.

Exceptional accuracy is achieved in forecasting, with correlation coefficients ranging from 96.9% to 97.2%.
The mean squared error and root mean squared error were utilized alongside other metrics for model evaluation during daytime predictions.
Analysis included comparing three neural network models: Bi-LSTM, GRU, and 1D-CNN, focusing on different scenarios with and without nighttime values used for training and prediction development. 1D-CNN showed accurate results in the second scenario but lagged behind GRU and Bi-LSTM models across both scenarios' evaluations.

Mansour et al. (2024) studied this question.

View Full Paper

Ask AI

Bookmark