Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Solar Irradiation Forecasting using Genetic Algorithms

V. Gunasekaran, K. K. Kovi, S. Arja, R. Chimata

TL;DR

The study addresses the challenge of accurate solar irradiation forecasting for grid management by comparing Linear Regression (LR), Extreme Gradient Boosting (XGB), and a Genetic Algorithm (GA) to optimize XGB hyperparameters using SURFRAD data from three U.S. stations. GA-based hyperparameter tuning automates model optimization and yields the highest predictive accuracy for Global Horizontal Irradiance (GHI) across diverse climates. Results show GA-optimized XGB achieving up to about 99% test accuracy with substantially lower MAE than LR or standard XGB, and around 97.75% validation accuracy with MAE near 7.45, highlighting the practical potential for operational forecasting. The work demonstrates the value of automated hyperparameter optimization in renewable-energy forecasting and suggests directions for scaling to more stations, incorporating additional features, and exploring ensemble approaches to further improve robustness and reliability.

Abstract

Renewable energy forecasting is attaining greater importance due to its constant increase in contribution to the electrical power grids. Solar energy is one of the most significant contributors to renewable energy and is dependent on solar irradiation. For the effective management of electrical power grids, forecasting models that predict solar irradiation, with high accuracy, are needed. In the current study, Machine Learning techniques such as Linear Regression, Extreme Gradient Boosting and Genetic Algorithm Optimization are used to forecast solar irradiation. The data used for training and validation is recorded from across three different geographical stations in the United States that are part of the SURFRAD network. A Global Horizontal Index (GHI) is predicted for the models built and compared. Genetic Algorithm Optimization is applied to XGB to further improve the accuracy of solar irradiation prediction.

Solar Irradiation Forecasting using Genetic Algorithms

TL;DR

The study addresses the challenge of accurate solar irradiation forecasting for grid management by comparing Linear Regression (LR), Extreme Gradient Boosting (XGB), and a Genetic Algorithm (GA) to optimize XGB hyperparameters using SURFRAD data from three U.S. stations. GA-based hyperparameter tuning automates model optimization and yields the highest predictive accuracy for Global Horizontal Irradiance (GHI) across diverse climates. Results show GA-optimized XGB achieving up to about 99% test accuracy with substantially lower MAE than LR or standard XGB, and around 97.75% validation accuracy with MAE near 7.45, highlighting the practical potential for operational forecasting. The work demonstrates the value of automated hyperparameter optimization in renewable-energy forecasting and suggests directions for scaling to more stations, incorporating additional features, and exploring ensemble approaches to further improve robustness and reliability.

Abstract

Renewable energy forecasting is attaining greater importance due to its constant increase in contribution to the electrical power grids. Solar energy is one of the most significant contributors to renewable energy and is dependent on solar irradiation. For the effective management of electrical power grids, forecasting models that predict solar irradiation, with high accuracy, are needed. In the current study, Machine Learning techniques such as Linear Regression, Extreme Gradient Boosting and Genetic Algorithm Optimization are used to forecast solar irradiation. The data used for training and validation is recorded from across three different geographical stations in the United States that are part of the SURFRAD network. A Global Horizontal Index (GHI) is predicted for the models built and compared. Genetic Algorithm Optimization is applied to XGB to further improve the accuracy of solar irradiation prediction.

Paper Structure

This paper contains 11 sections, 2 equations, 4 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Data Preprocessing
  3. Outlier Detection
  4. Feature Selection
  5. Machine Learning Algorithms
  6. Linear Regression (LR)
  7. Extreme Gradient Boosting (XGB)
  8. Genetic Algorithm (GA)
  9. Evaluation Metric
  10. Workflow Overview
  11. Conclusion

Figures (4)

  • Figure 1: Data preprocessing to obtain normalized distributions by removing outliers and cleaning the data. Each row shows the variable before (left) and after (right) preprocessing.
  • Figure 2: Feature importance used in the feature selection process (Random Forest). The top eight features are considered for the model.
  • Figure 3: Flow chart of the training and validation process.
  • Figure 4: Test (top) and validation (bottom) plots for the three stations depicting the datasets compared with the predicted values from LR, XGB, and GA models for Bondville, IL; Desert Rock, NV; and Penn State, PA during May.