"Comparative Analysis of Time-Series Supervised Learning Models for Predicting Solar and Wind Energy Outputs"

EasyChair Preprint 14449

Abstract

The growing global demand for renewable energy necessitates accurate and reliable forecasting methods to efficiently integrate solar and wind energy into power grids. This research focuses on a comparative analysis of various time-series supervised learning models for predicting solar and wind energy outputs. The study examines the performance of models such as Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Networks (CNN), and hybrid LSTM-CNN architectures, among others, to determine their efficacy in forecasting energy outputs under diverse climatic conditions.

The research begins with a comprehensive review of the characteristics of solar and wind energy, highlighting the inherent variability and challenges in prediction. It then delves into the selection criteria for supervised learning models, considering factors such as data requirements, computational complexity, and model interpretability. A robust methodology is developed, involving the collection of historical weather data and energy output from multiple geographically diverse sources. This data is used to train and validate the models, with a focus on optimizing hyperparameters to enhance prediction accuracy.

This study contributes to the advancement of predictive analytics in renewable energy, offering insights into the selection and optimization of time-series models for enhancing the reliability and efficiency of solar and wind energy forecasting.

Keyphrases: CNN (Convolutional Neural Networks), Energy Output Prediction, GRU (Gated Recurrent Unit), Hybrid LSTM-CNN, LSTM (Long Short-Term Memory), Real-time Data Assimilation, Solar Energy Prediction, Supervised Learning Models, Time Series Forecasting, Wind Energy Prediction, machine learning, renewable energy forecasting

@booklet{EasyChair:14449,
  author    = {Alakitan Samad},
  title     = {"Comparative Analysis of Time-Series Supervised Learning Models for Predicting Solar and Wind Energy Outputs"},
  howpublished = {EasyChair Preprint 14449},
  year      = {EasyChair, 2024}}