Application of Quantum Neural Network for Solar Irradiance Forecasting: A Case Study Using the Folsom Dataset, California DOI Open Access
Victor Oliveira Santos, Felipe Pinto Marinho, Paulo Alexandre Costa Rocha

et al.

Published: July 2, 2024

Quantum machine learning applications have become viable with the recent advancements in quantum computing. Merging ML power of computing holds great potential for data-driven decision-making, as well development more powerful models capable handling complex datasets faster processing time. This area offers improving accuracy real-time forecasting renewable energy production. However, literature on this topic is sparse. Addressing knowledge gap, study aims to design, implement, and evaluate performance a neural network forecast model solar irradiance up 3-hours ahead. The proposed was compared Support Vector Regression, Group Method Data Handling, Extreme Gradient Boost classical models. Using best configuration found, framework could provide competitive results when its competitors, considering intervals 5- 120-minutes ahead, where it fourth best-performing paradigm. For ahead predictions, QNN able overcome clas-sical counterparts, but XGBoost. fact can be an indication that may identify retrieve relevant spatiotemporal information from input dataset such manner not attainable by current approaches.

Language: Английский

Short-Medium-Term Solar Irradiance Forecasting with a CEEMDAN-CNN-ATT-LSTM Hybrid Model Using Meteorological Data DOI Creative Commons

M Mora Camacho,

Jorge Maldonado-Correa, Joel Torres-Cabrera

et al.

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(3), P. 1275 - 1275

Published: Jan. 26, 2025

In recent years, the adverse effects of climate change have increased rapidly worldwide, driving countries to transition clean energy sources such as solar and wind. However, these energies face challenges cloud cover, precipitation, wind speed, temperature, which introduce variability intermittency in power generation, making integration into interconnected grid difficult. To achieve this, we present a novel hybrid deep learning model, CEEMDAN-CNN-ATT-LSTM, for short- medium-term irradiance prediction. The model utilizes complete empirical ensemble modal decomposition with adaptive noise (CEEMDAN) extract intrinsic seasonal patterns irradiance. addition, it employs encoder-decoder framework that combines convolutional neural networks (CNN) capture spatial relationships between variables, an attention mechanism (ATT) identify long-term patterns, long short-term memory (LSTM) network dependencies time series data. This has been validated using meteorological data more than 2400 masl region characterized by complex climatic conditions south Ecuador. It was able predict at 1, 6, 12 h horizons, mean absolute error (MAE) 99.89 W/m2 winter 110.13 summer, outperforming reference methods this study. These results demonstrate our represents progress contributing scientific community field environments high its applicability real scenarios.

Language: Английский

Citations

1
Optimizing the Architecture of a Quantum–Classical Hybrid Machine Learning Model for Forecasting Ozone Concentrations: Air Quality Management Tool for Houston, Texas DOI Creative Commons
Victor Oliveira Santos, Paulo Alexandre Costa Rocha, Jesse Van Griensven Thé

et al.

Atmosphere, Journal Year: 2025, Volume and Issue: 16(3), P. 255 - 255

Published: Feb. 23, 2025

Keeping track of air quality is paramount to issue preemptive measures mitigate adversarial effects on the population. This study introduces a new quantum–classical approach, combining graph-based deep learning structure with quantum neural network predict ozone concentration up 6 h ahead. The proposed architecture utilized historical data from Houston, Texas, major urban area that frequently fails comply regulations. Our results revealed smoother transition between classical framework and its counterpart enhances model’s results. Moreover, we observed min–max normalization increased ansatz repetitions also improved hybrid performance. was evident evaluating assessment metrics root mean square error (RMSE), coefficient determination (R2) forecast skill (FS). Values for R2 FS horizons considered were 94.12% 31.01% 1 h, 83.94% 48.01% 3 75.62% 57.46% forecasts. A comparison existing literature both QML models methodology could provide competitive results, even surpass some well-established forecasting models, proving be valuable resource forecasting, thus validating this approach.

Language: Английский

Citations

0
Application of Quantum Neural Network for Solar Irradiance Forecasting: A Case Study Using the Folsom Dataset, California DOI Open Access
Victor Oliveira Santos, Felipe Pinto Marinho, Paulo Alexandre Costa Rocha

et al.

Published: July 2, 2024

Quantum machine learning applications have become viable with the recent advancements in quantum computing. Merging ML power of computing holds great potential for data-driven decision-making, as well development more powerful models capable handling complex datasets faster processing time. This area offers improving accuracy real-time forecasting renewable energy production. However, literature on this topic is sparse. Addressing knowledge gap, study aims to design, implement, and evaluate performance a neural network forecast model solar irradiance up 3-hours ahead. The proposed was compared Support Vector Regression, Group Method Data Handling, Extreme Gradient Boost classical models. Using best configuration found, framework could provide competitive results when its competitors, considering intervals 5- 120-minutes ahead, where it fourth best-performing paradigm. For ahead predictions, QNN able overcome clas-sical counterparts, but XGBoost. fact can be an indication that may identify retrieve relevant spatiotemporal information from input dataset such manner not attainable by current approaches.

Language: Английский

Citations

3