Optimizing Short-Term Photovoltaic Power Forecasting: A Novel Approach with Gaussian Process Regression and Bayesian Hyperparameter Tuning

Md. Samin Safayat Islam,Puja Ghosh,Md. Omer Faruque,Md. Rashidul Islam,Md. Alamgir Hossain,Md. Shafiul Alam,Md. Rafiqul Islam Sheikh

DOI: https://doi.org/10.3390/pr12030546

IF: 3.5

2024-03-11

Processes

Abstract:The inherent volatility of PV power introduces unpredictability to the power system, necessitating accurate forecasting of power generation. In this study, a machine learning (ML) model based on Gaussian process regression (GPR) for short-term PV power output forecasting is proposed. With its benefits in handling nonlinear relationships, estimating uncertainty, and generating probabilistic forecasts, GPR is an appropriate approach for addressing the problems caused by PV power generation's irregularity. Additionally, Bayesian optimization to identify optimal hyper-parameter combinations for the ML model is utilized. The research leverages solar radiation intensity data collected at 60-min and 30-min intervals over periods of 1 year and 6 months, respectively. Comparative analysis reveals that the data set with 60-min intervals performs slightly better than the 30-min intervals data set. The proposed GPR model, coupled with Bayesian optimization, demonstrates superior performance compared to contemporary ML models and traditional neural network models. This superiority is evident in 98% and 90% improvements in root mean square errors compared to feed-forward neural network and artificial neural network models, respectively. This research contributes to advancing accurate and efficient forecasting methods for PV power output, thereby enhancing the reliability and stability of power systems.

engineering, chemical

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the short - term prediction of photovoltaic (PV) power output. Specifically, since solar power generation is affected by meteorological variables such as solar radiation, cloud cover, humidity, pressure and temperature, its output is highly volatile and uncertain, which poses challenges to the stable operation of the power grid and the cost - effectiveness of the power system. Therefore, accurate prediction of PV power output is crucial for optimizing power dispatch, improving cost efficiency and ensuring reliable power supply. To address these challenges, the paper proposes a machine - learning model based on Gaussian process regression (GPR) and combines Bayesian optimization to adjust the hyper - parameters of the model in order to achieve more accurate short - term PV power prediction. By using the meteorological data of the Canberra region in Australia obtained from Solcast, the influence of data sets with different time intervals (60 minutes and 30 minutes) on the prediction performance was studied. The experimental results show that the proposed GPR model is significantly superior to the existing feed - forward neural network and artificial neural network models in prediction accuracy, especially in terms of root - mean - square error (RMSE), with improvements of 98% and 90% respectively. In summary, this research aims to improve the accuracy, robustness and flexibility of PV power prediction by developing a new prediction method, thereby enhancing the reliability and stability of the power system.

Optimizing Short-Term Photovoltaic Power Forecasting: A Novel Approach with Gaussian Process Regression and Bayesian Hyperparameter Tuning

Probabilistic Forecasting of Photovoltaic Generation: an Efficient Statistical Approach

Short-term photovoltaic power production forecasting based on novel hybrid data-driven models

Machine Learning Approaches for Short-Term Photovoltaic Power Forecasting

Enhanced solar photovoltaic power prediction using diverse machine learning algorithms with hyperparameter optimization

A review and evaluation of the state-of-the-art in PV solar power forecasting: Techniques and optimization

Enhancing PV power forecasting with deep learning and optimizing solar PV project performance with economic viability: A multi-case analysis of 10 MW Masdar project in UAE

Short-term prediction of photovoltaic power generation using Gaussian process regression

Enhancing photovoltaic energy forecasting: a progressive approach using wavelet packet decomposition

Short-term photovoltaic power prediction on modal reconstruction: A novel hybrid model approach

Improving Photovoltaic Power Prediction: Insights through Computational Modeling and Feature Selection

A short-term photovoltaic output power forecasting based on ensemble algorithms using hyperparameter optimization

Enhancing photovoltaic systems using Gaussian process regression for parameter identification and fault detection

Enhancing Short-Term Solar Photovoltaic Power Forecasting Using a Hybrid Deep Learning Approach

A Novel Improved Variational Mode Decomposition-Temporal Convolutional Network-Gated Recurrent Unit with Multi-Head Attention Mechanism for Enhanced Photovoltaic Power Forecasting

Forecasting of photovoltaic power generation and model optimization: A review

Optimal Scheduling of Energy Storage for Power System with Capability of Sensing Short-Term Future PV Power Production

A New Data Driven Long-Term Solar Yield Analysis Model of Photovoltaic Power Plants

Enhancing solar photovoltaic energy production prediction using diverse machine learning models tuned with the chimp optimization algorithm

Modeling optimization method based on Gamma test and NSGA II for forecast of PV power output

Short-term power prediction for photovoltaic power plants using a hybrid improved Kmeans-GRA-Elman model based on multivariate meteorological factors and historical power datasets