Tempo di lettura ca.: 1 minuti, 45 secondi
[[{“value”:”
Researchers have used the chimp optimization algorithm to optimize the hyperparameters of five PV power yield prediction machine learning models. This algorithm is based on the cooperative hunting behavior of chimpanzees in nature, mimicking the way they work together to target prey.
A scientific group led by researchers from German Jordanian University has analyzed the effect of the so-called chimp optimization algorithm (ChOA) on different PV yield production prediction machine learning (ML) models.
The ChOA is based on the cooperative hunting behavior of chimpanzees in nature, mimicking the way they work together to target prey, common amongst small mammals. They usually act in a group of three or four hunters and initially drive and block the prey, and then chase and attack them.
The algorithm explores different combinations of parameters to achieve the most promising result. It was used by the scientists to optimize the hyperparameters to five types of ML models. These include multiple linear regression (MLR), decision tree regression (DTR), random forest regression (RFR), support vector regression (SVR), and multi-layer perceptron (MLP).
“The effectiveness of this contribution is verified regarding data from a real case study, while resorting to various performance metrics from the literature including root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2),” the researchers explained.
Hyperparameters are external configurations set before the learning process begins that govern the learning process and do not change during training. Hyperparameters – such as learning rate in neural networks – influence the training dynamics and can, therefore, significantly impact the effectiveness of models.
All five models, with ChOA and without, were trained on 948 records and tested on 362 records. The records were taken between 2015 and 2018 from a 264 Kw PV system installed on a roof at the Applied Science University in Amman, the capital of Jordan. The installation tilt angle was set at 11 degrees and the azimuth angle to −36 degrees. Meteorological variables such as wind speed, relative humidity, ambient temperature, and solar irradiation were measured from a nearby weather station.
“Amman, Jordan, experiences a Mediterranean climate characterized by hot, dry summers and cool, wet winters,” the researchers added. “the average all-year temperature is 17.63 C, and the mean annual global horizontal irradiation stands at 2040.2 kWh/m2.”
Through this analysis, the scientists found that all models experienced performance improvements as a result of fine-tuning the hyperparameters using the ChOA.
“DTR exhibited substantial enhancements, with the testing RMSE decreasing to 1.972 and R2 increasing to 0.951,” they explained. “The RFR model showed notable improvements, with RMSE values decreasing to 1.773 for training and 1.837 for testing, and R2 values increasing to 0.964 for training and 0.963 for testing. The SVR model experienced the most remarkable enhancement, with the testing RMSE dropping to 0.818 and R2 increasing to 0.977.”
Post-ChOA optimization, MLP was found to show the best results in predicting PV power yield. Specifically, it was able to reach 0.503, 0.397, and 0.99 in RMSE, MAE, and R2, respectively. “The ChOA effectively fine-tuned the parameters, resulting in improved model fitting, reduced overfitting, and enhanced generalization compared to two other widely used optimization algorithms from the literature: particle swarm optimization (PSO) and genetic algorithm (GA),” the team concluded.
The results were presented in “Enhancing solar photovoltaic energy production prediction using diverse machine learning models tuned with the chimp optimization algorithm,” published in Scientific Reports. The group included academics from Jordan’s German Jordanian University, University of Jordan, Al-Balqa Applied University, and Alabama’s Tuskegee University.
“}]]