Valorisez vos réalisations phares sur People@HES-SO Plus d'infos
PEOPLE@HES-SO – Annuaire et Répertoire des compétences
PEOPLE@HES-SO – Annuaire et Répertoire des compétences

PEOPLE@HES-SO
Annuaire et Répertoire des compétences

Aide
language
  • fr
  • en
  • de
  • fr
  • en
  • de
  • SWITCH edu-ID
  • Administration
ID
« Retour
Jafari Fereshteh

Jafari Fereshteh

Doctorant-e

Compétences principales

Control Engineering

Renewable Energy | Solar Panels

Data-Driven Modeling

Machine Learning

Photovoltaic

Hybrid Solar Panels

PV Power Prediction

  • Contact

  • Recherche

  • Publications

Contrat principal

Doctorant-e

Bureau: ENP.23.N412

HES-SO Valais-Wallis - Haute Ecole d'Ingénierie
Rue de l'Industrie 23, 1950 Sion, CH
HEI - VS

A determined, diligent, and motivated engineer in Electrical Engineering, who works on photovoltaic systems (PV, PV-T, CPV-T) using analytical, data-driven, and machine-learning algorithms for modeling and control; I have the ability to learn new theoretical methods and technical software quickly and well.

 Currently, I am pursuing a PhD at the University of Bern and HES-SO in Switzerland as part of a European project: Increasing Control and Efficiency in Regional Energy Systems Through Quantum Sensors and Machine Learning. My research focuses on photovoltaic (PV) power prediction, energy system optimization, and integrating machine learning with power networks. Passionate about teaching and sharing knowledge, I am committed to advancing sustainable energy solutions through innovation and collaboration.

I am very interested in welcoming challenging situations at laboratories and off-shore /on-shore sites. I am usually a group leader who loves to work as an effective team member with commitment, responsibility, and special sensitivity to timeliness. I try not to be a one-dimensional girl but also a good and influential person for this planet and its great people.

Aucune donnée à afficher pour cette section

En cours

Increasing Control and Efficiency in Regional Energy Systems Through Quantum Sensors and Machine Learning

Rôle: Collaborateur/trice

Financement: SNSF

Description du projet :

QuantumIRES aims to advance secure, resilient, and carbon-free regional energy systems while supporting inter-regional energy exchange and European energy stability. Integrating cutting-edge research in time series modeling, sky cameras, image processing, Quantum sensors, and AI-driven solutions fosters collaboration among researchers and solution providers to automate and optimize energy management.

Equipe de recherche au sein de la HES-SO: Moerschell Joseph , Jafari Fereshteh , Praplan Charles

Statut: En cours

Terminés

Solar Energy Optimization Via a Remote Research Platform

Rôle: Collaborateur/trice

Financement: SNSF

Description du projet :

This project focuses on hybrid solar technologies, including PV, PV-T, and CPV-T systems with concentrating mirrors. It combines analytical and data-driven modeling to enhance performance analysis and optimization. The work also involves generating datasets through real measurements to bridge the gap between theoretical insights and practical applications.

 

Equipe de recherche au sein de la HES-SO: Moghaddam Fariba , Jafari Fereshteh , Vaccari Aldo , Bonvin Axel

Partenaires académiques: Qobad Shafiee, University of Kurdistan; Kamran Moradi, University of Kurdistan; Léo Agrusti, HES-SO; Valérie Pompili, HES-SO

Durée du projet: 01.09.2023 - 31.08.2024

Publications liées:

  • Data-Driven Performance Modeling of Solar Panels Using Polynomial Regression
  • Modeling of Photovoltaic-Thermal Systems Using Multivariate Polynomial Regression

Statut: Terminé

2025

Data-Driven Modeling of Photovoltaic Panels Using Neural Networks
Article scientifique

Jafari Fereshteh, Kamran Moradi, Moghaddam Fariba, Qobad Shafiee

IEEE, 2025 , pp.  326-333

Lien vers la publication

Résumé:

Photovoltaic (PV) systems are essential for the shift towards sustainable energy. Accurate performance modeling of these systems is vital for optimizing their efficiency and adapting to changing environmental conditions. Traditional analytical models often struggle with real-world complexities. This paper uses a neural networks (NNs) method to model PV system performance, offering superior accuracy and adaptability. By leveraging NNs, the model adopts a black-box methodology, bypassing the need for detailed knowledge of PV system complexities. The study utilizes historical performance data from two panels located in Sion, Switzerland, and Eugene, Oregon, in the United States. This approach captures the interdependencies between input parameters such as solar irradiance, temperature, and applied load and output variables such as voltage and current. Additionally, it accounts for key PV metrics including open circuit voltage, short circuit current, maximum power point current, and maximum power point voltage. The NN model is trained with a comprehensive dataset featuring high persistent excitation along with real measurements. Performance metrics such as root mean square error demonstrate that the data-driven NN model surpasses traditional methods, offering a more practical solution for PV system performance modeling.

2024

Modeling of Photovoltaic-Thermal Systems Using Multivariate Polynomial Regression
Article scientifique

Kamran Moradi, Jafari Fereshteh, Moghaddam Fariba, Qobad Shafiee

3rd IFAC Workshop on Integrated Assessment Modeling for Environmental Systems IAMES 2024: Savona, Italy, May 29 – May 31, 2024, 2024

Lien vers la publication

Résumé:

A challenge prevalent in the photovoltaic (PV) system applications in the domain of control engineering lies in the formulation and refinement of their models. Among the esteemed configurations in this domain, the photovoltaic-thermal (PVT) system emerges as particularly noteworthy. Furthermore, the appeal of employing a modeling approach with minimal complexity is noteworthy in this context. To achieve this objective, there is a growing trend in utilizing machine learning (ML) approaches, known for their data-driven modeling capabilities and minimal complexity. In this paper, the application of the multivariate polynomial regression, a straightforward ML method, has been employed to model the PVT system. This choice aims to address the challenge of model complexity inherent in conventional mathematical method, enabling its utility for control purposes. The proposed method is implemented using empirical data derived from laboratory PVT setups, incorporating geographic inputs situated in Sion, Switzerland. A comprehensive comparison with the conventional method has been undertaken for evaluation purposes. The results, encompassing electrical and thermal powers of the PVT system, indicate that the proposed method achieves significantly greater accuracy compared to the conventional.

Data-Driven Performance Modeling of Solar Panels Using Polynomial Regression
Article scientifique

Jafari Fereshteh, Kamran Moradi, Qobad Shafiee, Moghaddam Fariba

2024 International Conference on Control, Automation and Diagnosis (ICCAD), 2024

Lien vers la publication

Résumé:

Photovoltaic (PV) systems are integral to renewable energy, demanding accurate performance modeling for optimal functionality. This paper presents a pragmatic, data-driven approach employing Polynomial Regression (PR) for solar panel modeling to boost accuracy and adaptability to environmental variables. By emphasizing the advantages of data-driven models in attaining predictive precision, this study aims to reconcile theoretical concepts with practical solar panel performance. PR is employed as a black-box machine learning (ML) algorithm to transcend the limitations of traditional modeling methods, revealing ML's ability to grasp intricate relationships and ensure precise predictions. Utilizing real and simulated datasets encompassing solar irradiance, ambient temperature, and applied load through hardware in the loop (HIL), the model is trained to forecast the electrical outputs of solar panels in two approaches to estimate output voltage and current as well as key points on the panels' I-V curve. Assessment of model accuracy using Root Mean Square Error (RMSE) showcases the PR model’s capacity to accurately depict solar panel performance by capturing non-linear relationships between environmental variable conditions and electrical outputs. Experimentation validates the method's effectiveness, focusing on accuracy, generalization to testing, and on-site data validation in two approaches. This research endeavors to bridge the gap between theory and practice, propelling advancements in the field of solar PV systems and facilitating more efficient solar energy utilization.

Shallow Learning vs. Deep Learning in Engineering Applications
Chapitre de livre

Jafari Fereshteh, Kamran Moradi, Qobad Shafiee

,  Shallow Learning vs. Deep Learning: A Practical Guide for Machine Learning Solutions. 2024,  Switzerland : Springer Nature Switzerland

Lien vers la publication

Résumé:

In this chapter, the application of machine learning (ML) in various engineering domains has ushered in transformative advancements, offering solutions to intricate problems and paving the way for unprecedented innovation. Shallow learning methods, known for their simplicity and interpretability, prove effective in scenarios with limited datasets, computational constraints, and straightforward variable relationships. Ideal for engineers seeking practical solutions in challenging data collection environments, these methods excel with smaller datasets. Conversely, deep learning has expanded ML applications by extracting intricate patterns from vast datasets, offering unparalleled accuracy in complex variable relationships. However, this comes at the expense of increased computational requirements, substantial labeled data needs, and reduced interpretability due to higher complexity. In evolving engineering applications, the selection between shallow and deep learning methodologies depends on specific requirements, constraints, and the nature of the problem, requiring practitioners to carefully assess the advantages and disadvantages for optimal utilization of ML in innovative solutions. In this chapter, an exploration of these methodologies has been undertaken across diverse fields of engineering. Additionally, an illustrative example within the realm of electrical engineering has been presented to facilitate a comparative analysis between shallow learning and deep learning methods.

2021

Investigation of Offshore Wind Turbine Foundation - Floating Offshore and Fixed Base Offshore - and Potential of North and South Seas of Iran
Article scientifique

Jafari Fereshteh, Hajieh Bastami

Karafan Journal, 2021 , vol.  18, no  3, pp.  207-235

Lien vers la publication

Résumé:

In the present work, the potential and capability of Iran's waters for acquiring and installing different types of wind turbines is discussed. For this purpose, various types of offshore wind turbines such as offshore float as the latest type of wind turbines as well as various types of base fixed offshore wind turbines were studied. In addition to reviewing the state of wind turbine technology in Iran according to oceanographic features, geographical coordinates, transportation costs and the possibility of assembly and ease of installation for the northern and southern seas of Iran, the amount of wind resources, water depth and available area were also taken into consideration. Based on the SWOT model, recommendations were made for the use of different types of turbines. In a part of the research, a comparison was made between Iran and leading countries in this field. For the Caspian Sea, the type of traction base of these turbines is not recommended due to the structural complexity, difficult installation and transportation and the need for deep water, while monopile structures and the third generation of gravity are suitable options for the shallow depths of lakes. In addition, near the southern coast of Iran, due to high tides and many changes in water depth during the day and night and the impact of the tension of the restraint lines of traction platforms, the use of this structure is not recommended. Due to the high depth of water in coastal areas and very good stability in various weather conditions, spar structures are introduced as a desirable option.

Réalisations

Médias et communication
Nous contacter
Suivez la HES-SO
linkedin instagram facebook twitter youtube rss
univ-unita.eu www.eua.be swissuniversities.ch
Mentions légales
© 2021 - HES-SO.

HES-SO Rectorat