A Comparative Study on Data Augmentation Using Generative Models for Robust Solar Irradiance Prediction
1
조회
0
다운로드
3
인용
0
참고문헌
키워드
초록
In this paper, we propose a method to enhance the prediction accuracy of solar irradiance for three major South Korean cities: Seoul, Busan, and Incheon. Our method entails the development of five generative models—vanilla GAN, CTGAN, Copula GAN, WGANGP, and TVAE—to generate independent variables that mimic the patterns of existing training data. To mitigate the bias in model training, we derive values for the dependent variables using random forests and deep neural networks, enriching the training datasets. These datasets are integrated with existing data to form comprehensive solar irradiance prediction models. The experimentation revealed that the augmented datasets led to significantly improved model performance compared to those trained solely on the original data. Specifically, CTGAN showed outstanding results due to its sophisticated mechanism for handling the intricacies of multivariate data relationships, ensuring that the generated data are diverse and closely aligned with the real-world variability of solar irradiance. The proposed method is expected to address the issue of data scarcity by augmenting the training data with high-quality synthetic data, thereby contributing to the operation of solar power systems for sustainable development.
Abstract
In this paper, we propose a method to enhance the prediction accuracy of solar irradiance for three major South Korean cities: Seoul, Busan, and Incheon. Our method entails the development of five generative models—vanilla GAN, CTGAN, Copula GAN, WGANGP, and TVAE—to generate independent variables that mimic the patterns of existing training data. To mitigate the bias in model training, we derive values for the dependent variables using random forests and deep neural networks, enriching the training datasets. These datasets are integrated with existing data to form comprehensive solar irradiance prediction models. The experimentation revealed that the augmented datasets led to significantly improved model performance compared to those trained solely on the original data. Specifically, CTGAN showed outstanding results due to its sophisticated mechanism for handling the intricacies of multivariate data relationships, ensuring that the generated data are diverse and closely aligned with the real-world variability of solar irradiance. The proposed method is expected to address the issue of data scarcity by augmenting the training data with high-quality synthetic data, thereby contributing to the operation of solar power systems for sustainable development.
참고문헌 (31)
- 한유나, "강건하고 설명 가능한 스마트 미터 데이터 결측치 대체 기법", 한국전자거래학회지, 27(3), pp.21~43 (2022), 한국전자거래학회
- L. Cozzi, "World energy outlook 2020", International Energy Agency, 1~461 (2020)
- "2023 SEMI-ANNAL Report"[URL]
- Abdeen Mustafa Omer, "Energy, environment and sustainable development", Renewable and Sustainable Energy Reviews, 12(9), pp.2265~2300 (2008), Elsevier BV[DOI]
- 문지훈, "Multistep-Ahead Solar Irradiance Forecasting for Smart Cities Based on LSTM, Bi-LSTM, and GRU Neural Networks", 한국전자거래학회지, 27(4), pp.27~52 (2022), 한국전자거래학회
- Can Wan, "Photovoltaic and solar power forecasting for smart grid energy management", CSEE Journal of Power and Energy Systems, 1(4), pp.38~46 (2015), Power System Technology Press[DOI]
- Jinwoong Park, "Multistep-Ahead Solar Radiation Forecasting Scheme Based on the Light Gradient Boosting Machine: A Case Study of Jeju Island", Remote Sensing, 12(14), pp.2271~ (2020), MDPI AG[DOI]
- Dayeong So, "A Hybrid Ensemble Model for Solar Irradiance Forecasting: Advancing Digital Models for Smart Island Realization", Electronics, 12(12), pp.2607~ (2023), MDPI AG[DOI]
- Cyril Voyant, "Machine learning methods for solar radiation forecasting: A review", Renewable Energy, 105, pp.569~582 (2017), Elsevier BV[DOI]
- Dayeong So, "BiGTA-Net: A Hybrid Deep Learning-Based Electrical Energy Forecasting Model for Building Energy Management Systems", Systems, 11(9), pp.456~ (2023), MDPI AG[DOI]
- J. Lee, "Reliable solar irradiance prediction using ensemble learning-based models: A comparative study", Energy Conversion and Management, 208, pp.112582~ (2020)[DOI]
- Pratima Kumari, "Deep learning models for solar irradiance forecasting: A comprehensive review", Journal of Cleaner Production, 318, pp.128566~ (2021), Elsevier BV[DOI]
- Ke Yan, "Short-Term Solar Irradiance Forecasting Based on a Hybrid Deep Learning Methodology", Information, 11(1), pp.32~ (2020), MDPI AG[DOI]
- Subeen Leem, "Enhancing multistep-ahead bike-sharing demand prediction with a two-stage online learning-based time-series model: insight from Seoul", The Journal of Supercomputing (2023), Springer Science and Business Media LLC[DOI]
- Jie Gui, "A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications", IEEE Transactions on Knowledge and Data Engineering, 35(4), pp.3313~3332 (2023), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Xiaoqiao Huang, "Time series forecasting for hourly photovoltaic power using conditional generative adversarial network and Bi-LSTM", Energy, 246, pp.123403~ (2022), Elsevier BV[DOI]
- Qianqian Li, "A Solar Irradiance Forecasting Framework Based on the CEE-WGAN-LSTM Model", Sensors, 23(5), pp.2799~ (2023), MDPI AG[DOI]
- Jaeuk Moon, "Conditional Tabular GAN-Based Two-Stage Data Generation Scheme for Short-Term Load Forecasting", IEEE Access, 8, pp.205327~205339 (2020), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Fei Wang, "Generative adversarial networks and convolutional neural networks based weather classification model for day ahead short-term photovoltaic power forecasting", Energy Conversion and Management, 181, pp.443~462 (2019), Elsevier BV[DOI]
- Mohammad Navid Fekri, "Generating Energy Data for Machine Learning with Recurrent Generative Adversarial Networks", Energies, 13(1), pp.130~ (2019), MDPI AG[DOI]
- Chenlu Tian, "Data driven parallel prediction of building energy consumption using generative adversarial nets", Energy and Buildings, 186, pp.230~243 (2019), Elsevier BV[DOI]
- L. Xu, "Modeling tabular data using conditional GAN", Advances in Neural Information Processing Systems, 32 (2019)[DOI]
- M. Arjovsky, "Wasserstein Generative Adversarial Networks", Proceedings of the 34th International Conference on Machine Learning (2017)
- H. Ishfaq, "TVAE: Deep metric learning approach for variational autoencoder", roceedings of the Workshop at International Conference on Learning Representations, 32 (2018)
- J. Moon, "A comparative analysis of tree-based models for day-ahead solar irradiance forecasting", Proceedings of the 2021 International Conference on Platform Technology and Service, pp.1~6 (2021)
- Leo Breiman, "Random forests", Machine Learning, 45(1), pp.5~32 (2001), Springer Science and Business Media LLC[DOI]
- Joohyun Jang, "RAID: Robust and Interpretable Daily Peak Load Forecasting via Multiple Deep Neural Networks and Shapley Values", Sustainability, 15(8), pp.6951~ (2023), MDPI AG[DOI]
- V. Borisov, "Deep neural networks and tabular data: A survey", IEEE Transactions on Neural Networks for Learning Systems, pp.1~21 (2022)
- Jihoon Moon, "Hybrid Short-Term Load Forecasting Scheme Using Random Forest and Multilayer Perceptron", Energies, 11(12), pp.3283~ (2018), MDPI AG[DOI]
- "Orange Data Mining"[URL]