Article
Green & Sustainable Science & Technology
Can Wan, Weiting Qian, Changfei Zhao, Yonghua Song, Guangya Yang
Summary: This paper proposes a probabilistic forecasting-based HESS sizing and control scheme to cost-effectively smooth wind power fluctuations, which significantly reduces the installation cost and operation cost of HESS and prominently smooth the wind power fluctuations, as demonstrated by case studies based on actual data from a Danish wind farm.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY
(2021)
Article
Energy & Fuels
Jennie Molinder, Sebastian Scher, Erik Nilsson, Heiner Kornich, Hans Bergstrom, Anna Sjoblom
Summary: A probabilistic machine learning method, quantile regression forests, is applied to forecast icing-related production losses in wind energy in cold climates. The method has shown to produce valuable probabilistic forecasts and the output from a physical icing model enhances forecast skill when combined with Numerical Weather Prediction data. Training data from other icing-affected stations can also increase the forecast accuracy by providing additional data which poses a challenge in forecasting for wind energy in cold climates.
Article
Green & Sustainable Science & Technology
Lars odegaard Bentsen, Narada Dilp Warakagoda, Roy Stenbro, Paal Engelstad
Summary: This study investigates uncertainty modeling in wind power forecasting using different parametric and non-parametric methods. Johnson's SU distribution is found to outperform Gaussian distributions in predicting wind power. This research contributes to the literature by introducing Johnson's SU distribution as a candidate for probabilistic wind forecasting.
JOURNAL OF CLEANER PRODUCTION
(2024)
Article
Thermodynamics
Huijing Fan, Zhao Zhen, Nian Liu, Yiqian Sun, Xiqiang Chang, Yu Li, Fei Wang, Zengqiang Mi
Summary: This paper proposes a novel probabilistic forecasting method based on SDA, FCM, LSTM, and KDE, considering the correlation between wind power fluctuation patterns and forecasting errors. Simulation results show that introducing pattern recognition can improve the skill score of probabilistic forecasting by 36.50% on average.
Article
Engineering, Electrical & Electronic
Yunyi Li, Can Wan, Zhaojing Cao, Yonghua Song
Summary: This paper proposes a novel data-driven integrated probabilistic forecasting and analysis methodology for the nonparametric probabilistic optimal power flow (N-POPF), which utilizes uncertainty analysis to guide wind power forecasting and analysis. The proposed methodology is validated to be superior in estimation accuracy and computational efficiency.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2023)
Article
Energy & Fuels
Jiani Heng, Yongmiao Hong, Jianming Hu, Shouyang Wang
Summary: This study aims to improve probabilistic wind speed forecasting by using characteristic information from wind farms, evaluating various probability density functions, and constructing Generalized Gaussian Process models. A pseudo-likelihood method is proposed to enhance model robustness, ultimately leading to improved accuracy in probabilistic wind speed forecasting.
Article
Green & Sustainable Science & Technology
Leandro Von Krannichfeldt, Yi Wang, Thierry Zufferey, Gabriela Hug
Summary: This paper studies an online ensemble approach for probabilistic wind power forecasting by leveraging the most recent information and combining multiple forecasting models, improving the forecasting performance.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY
(2022)
Article
Thermodynamics
Xiaochong Dong, Yingyun Sun, Lei Dong, Jian Li, Yan Li, Lei Di
Summary: This paper proposes a multi-domain adversarial network (MDAN) to address the challenge of wind power forecasting for newly-built wind farms due to limited historical data availability. The MDAN utilizes multi-domain datasets and a data fusion-based incremental learning method to reduce domain shift and mitigate catastrophic forgetting. The results show that MDAN provides accurate short-term wind power probabilistic forecasts in zero-shot learning and enhances forecast accuracy in few-shot learning. Visualization analysis confirms the success of MDAN in reducing domain shift between source and target domains.
Article
Thermodynamics
Weichao Dong, Hexu Sun, Jianxin Tan, Zheng Li, Jingxuan Zhang, Huifang Yang
Summary: This paper proposes a regional wind power probabilistic forecasting model based on IKDE, regular vine copulas, and ensemble learning, aiming to improve the prediction accuracy of wind power generation through optimizing marginal PDF and joint distribution functions. In addition, the introduction of the MD-MTD method for prediction improvement with insufficient data has been validated, demonstrating that the proposed model performs well in wind power generation prediction.
Article
Engineering, Electrical & Electronic
Evelyn Heylen, Jethro Browell, Fei Teng
Summary: This paper develops a model to produce calibrated, data-driven probabilistic forecasts of the inertia contribution of transmission-connected synchronous generators. The model quantifies forecast uncertainty, allowing system operators to manage the risk of frequency instability cost-effectively. The paper also emphasizes the importance of adopting a rolling horizon forecast approach to deal with the rapidly changing characteristics of the inertial response in power systems.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2022)
Article
Thermodynamics
Shilin Sun, Yuekai Liu, Qi Li, Tianyang Wang, Fulei Chu
Summary: This paper proposes a novel method to enhance the reliability of wind condition knowledge by considering the spatial information of surrounding wind turbines and achieve wind power modeling using transformer neural networks based on the multi-head attention mechanism. Experimental results show that the proposed method outperforms other approaches, especially in large steps forecasting, with significantly better average values of mean absolute error.
ENERGY CONVERSION AND MANAGEMENT
(2023)
Article
Green & Sustainable Science & Technology
Mahdieh Shamsi, Paul Cuffe
Summary: This paper demonstrates how a binary prediction market can be used to achieve reliable renewable energy forecasts. By trading shares, the probability of the outcomes can be determined, and through regression analysis, the complete distribution function of renewable energy output can be extracted. The proposed method performs well in test cases in Australia, reducing electricity market imbalance costs.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY
(2022)
Article
Green & Sustainable Science & Technology
Honglin Wen, Jinghuan Ma, Jie Gu, Lyuzerui Yuan, Zhijian Jin
Summary: In this paper, a probabilistic wind power forecasting model is proposed through quantification of epistemic uncertainty and aleatory uncertainty. The use of sparse variational Gaussian process helps to address inference complexity and hyperparameters determination issues, resulting in a model that is comparable to state-of-the-art in terms of continuous ranked probability score.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY
(2022)
Article
Energy & Fuels
Yanli Liu, Junyi Wang
Summary: This paper proposes a transfer learning-based method for probabilistic wind power forecasting. It utilizes model-based transfer learning to construct a multilayer extreme learning machine, optimizes the output mapping factors using particle swarm optimization, and updates the weights through joint distribution adaptation. The method achieves more accurate quantile forecasting results and better nonlinear fitting ability compared to other methods.
Article
Economics
Ciaran Gilbert, Jethro Browell, David McMillan
Summary: This study introduces a novel method for predicting safety-critical access conditions by generating density forecasts of sea-state variables and using them as inputs to a vessel motion model. This approach improves the safety and efficiency of crew transfers.
INTERNATIONAL JOURNAL OF FORECASTING
(2021)
Article
Energy & Fuels
Lubos Buzna, Pasquale De Falco, Gabriella Ferruzzi, Shahab Khormali, Daniela Proto, Nazir Refa, Milan Straka, Gijs van der Poel
Summary: This paper presents a methodology for probabilistic electric vehicle load forecasting for different geographic regions, using a hierarchical approach to decompose the problem at low-level regions and forecast the aggregate load at a high-level geographic region through an ensemble methodology. Experimental results show that hierarchical approaches increase the skill of probabilistic forecasts up to 9.5% compared with non-hierarchical approaches.
Article
Chemistry, Multidisciplinary
Pasquale De Falco, Pietro Varilone
Summary: This paper studies the statistical characterization of supraharmonics in low-voltage distribution networks, proposing several probability distributions and discussing how multimodal distributions can fit more general scenarios where supraharmonic emissions follow regime patterns. Numerical experiments based on actual data suggest that multimodal distributions are useful in characterizing supraharmonics in most cases, even in the presence of stair-shaped empirical distributions. This research can serve as a starting point for the development of probabilistic power system analysis tools considering supraharmonic emissions and striving towards standardization in the 2-150 kHz range.
APPLIED SCIENCES-BASEL
(2021)
Editorial Material
Computer Science, Information Systems
Elio Chiodo, Pasquale De Falco, Luigi Pio Di Noia
Article
Engineering, Electrical & Electronic
Antonio Bracale, Pierluigi Caramia, Guido Carpinelli, Pasquale De Falco
Summary: This paper presents the novel comprehensive probabilistic tool SmarTrafo, which predicts the probability of success in operating transformers by DTR through an exact analytic stress-strength model, and formulates an alarm-setting strategy to establish warnings of potential risks. The proposal is confirmed to be suitable in predicting the probability of success and establishing high-performance alarms based on actual data experiments.
IEEE TRANSACTIONS ON POWER DELIVERY
(2021)
Article
Energy & Fuels
Mokhtar Bozorg, Antonio Bracale, Mauro Carpita, Pasquale de Falco, Fabio Mottola, Daniela Proto
Summary: Modern distribution systems are increasingly incorporating photovoltaic generation systems due to the uncertain nature of solar primary sources. Developing photovoltaic power forecasting models is essential for smart distribution networks. Probabilistic forecasts provide additional flexibility for decision making and optimization strategies in energy management systems.
Article
Engineering, Multidisciplinary
Pasquale De Falco, Luigi Pio Di Noia, Renato Rizzo
Summary: Lithium-ion batteries, as the core equipment of electric vehicles, are expected to spread out through smart power systems and are subject to degradation over time, which requires battery prognostic to track the degradation and schedule maintenance and replacement before failures. In this article, multiple linear regression models are developed to predict battery SoH based on historical charge/discharge data and validated through numerical experiments based on actual data.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2021)
Article
Engineering, Electrical & Electronic
Antonio Bracale, P. Caramia, P. De Falco, A. R. Di Fazio, P. Varilone
Summary: The analysis of active unbalanced three-phase 4-wire distribution networks under short-circuit conditions has become important due to the increasing diffusion of single-phase Photovoltaic Systems. The main challenge lies in modeling the complex behavior of the inverter control systems accurately.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2022)
Article
Engineering, Multidisciplinary
Elio Chiodo, Pasquale De Falco, Luigi Pio Di Noia
Summary: In this article, a hybrid methodology is developed to characterize the remaining useful life (RUL) of lithium-ion batteries even with limited data. The methodology considers several probability distributions and uses expectation-maximization algorithms for parameter estimation. The proposed approach is tested on a RUL dataset created by Monte Carlo sampling on an electrochemical battery model. Numerical experiments are reported to evaluate the effectiveness of the proposal.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2022)
Article
Engineering, Multidisciplinary
Antonio Bracale, Pasquale De Falco, Luigi Pio Di Noia, Renato Rizzo
Summary: In this paper, probabilistic models based on time series and regression methods are developed and compared for predicting the remaining useful life of lithium-ion batteries. The models are developed using data from accelerated degradation tests and different approaches are used for predicting battery health and lifespan. The results demonstrate the accuracy and effectiveness of the proposed models.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2023)
Proceedings Paper
Green & Sustainable Science & Technology
Andrea Altomonte, Antonio Bracale, Pierluigi Caramia, Pasquale De Falco, Giovanni Di Ilio, Luigi Pio Di Noia, Elio Jannelli, Renato Rizzo
Summary: This paper analyzes and simulates the power unit of a hybrid fuel cell/battery truck with an electric motor, aiming to implement cleaner technology in the propulsion of heavy-duty vehicles and reduce the environmental impact of the port-logistic industry.
2023 IEEE IAS GLOBAL CONFERENCE ON RENEWABLE ENERGY AND HYDROGEN TECHNOLOGIES, GLOBCONHT
(2023)
Proceedings Paper
Energy & Fuels
Antonio Bracale, Pierluigi Caramia, Guido Carpinelli, Pasquale De Falco, Angela Russo
Summary: Dynamic Transformer Rating (DTR) is a method for optimizing transformer performance under time-varying load and/or environmental conditions. Recent studies suggest that DTR should consider limitations on current, Hottest-Spot Temperature (HST), and Top-Oil Temperature (TOT) simultaneously. This paper proposes a novel probabilistic approach to predict the probability of these limits being exceeded and provide warnings if the probability is below a threshold.
2022 17TH INTERNATIONAL CONFERENCE ON PROBABILISTIC METHODS APPLIED TO POWER SYSTEMS (PMAPS)
(2022)
Proceedings Paper
Engineering, Electrical & Electronic
Mohammad Rayati, Mokhtar Bozorg, Mauro Carpita, Pasquale De Falco, Pierluigi Caramia, Antonio Bracale, Daniela Proto, Fabio Mottola
Summary: This paper evaluates the effectiveness of forecasting and optimization algorithms on a laboratory platform that mimics a domestic distribution grid with a high penetration of photovoltaic systems. The algorithms ensure efficient and secure real-time operation of the grid and the provision of flexibility services. Uncertainties arise from variations in power production and consumption, as well as real-time deployment of flexibility services.
2022 IEEE 21ST MEDITERRANEAN ELECTROTECHNICAL CONFERENCE (IEEE MELECON 2022)
(2022)
Proceedings Paper
Energy & Fuels
Antonio Bracale, Pierluigi Caramia, Pasquale De Falco, Enrica Di Mambro, Pietro Varilone, Paola Verde
Summary: This paper analyzes the impact of Synchronous Compensators (SyCs) on the robustness of networks with a high share of Renewable Energy Sources (RES) during short-circuit conditions. The utility of different SyC installations and configurations is investigated through a real case study using robustness indices tailored for this purpose.
2022 20TH INTERNATIONAL CONFERENCE ON HARMONICS & QUALITY OF POWER (ICHQP 2022)
(2022)
Proceedings Paper
Energy & Fuels
Antonio Bracale, Pierluigi Caramia, Pasquale De Falco, Guido Carpinelli
Summary: This paper investigates probabilistic forecasting methods for predicting waveform distortion indices. The results show that using the quantile regression model and principal component analysis can effectively predict waveform distortion indices in power systems.
2022 20TH INTERNATIONAL CONFERENCE ON HARMONICS & QUALITY OF POWER (ICHQP 2022)
(2022)
Article
Computer Science, Information Systems
Guido Carpinelli, Antonio Bracale, Pietro Varilone, Tomasz Sikorski, Pawel Kostyla, Zbigniew Leonowicz
Summary: This paper discusses a joint method for improving the performance of existing methods in evaluating waveform distortion in Smart Grids. The method utilizes Discrete Wavelet Transform to divide the waveform, and then combines the sliding-window modified ESPRIT method and sliding-window Discrete Fourier Transform to analyze low and high-frequency bands, respectively, in order to enhance the accuracy of estimating each frequency spectral component.