We combine data-driven models with power electronics physics to build tools that run in real time. Our work includes fault detection and localization in converters with a high number of switches (CNN, LSTM and convolutional-transformer networks), physics-informed and LSTM-based grid impedance and parameter estimation, digital twins for condition monitoring and predictive maintenance, and neural-network-enhanced predictive control.
Key capabilities
- Switch fault detection and localization
- Grid impedance and parameter estimation (PINN, LSTM)
- Digital twins and predictive maintenance
- AI-enhanced and data-driven control
Selected publications
- A Novel Compact Hierarchical Deep Convolutional Neural Network Architecture for Faulty Switch Detection and Localization in Power Converters
IEEE Transactions on Industry Applications, 2026 - A Novel Convolutional-Transformer Hybrid Neural Network Framework for Switch Fault Diagnosis in Power Converters
IEEE Transactions on Industry Applications, 2026 - Fault Diagnosis Using Shallow Neural Networks for Voltage Source Inverters in Motor Drives
IEEE Transactions on Industry Applications, 2024 - A Data-Driven Fault Diagnostics Approach for Dual Active Bridge Converters
IEEE IAS Annual Meeting, 2025 - ANN-Enhanced Modulated Model Predictive Control for AC-DC Converters in Grid-Connected Battery Systems
Energies, 2025 - A Novel Grid Impedance Estimation Method for Grid-Connected Inverter Systems Using LSTM Neural Networks
IEEE ECCE, 2025
Team
Prof. Pericle Zanchetta
Full Professor
Dr. Giulia Tresca
Assistant Professor
Simone Cossu
PhD Student
Farnoush Shamsazad
PhD Student
← Back to Research · All publications
Work with us
To discuss a collaboration, a project idea or a technical challenge in this area, contact:
Prof. Pericle Zanchetta
Full Professor – PELAB Director
pericle.zanchetta@unipv.it
Dr. Giulia Tresca
Assistant Professor
giulia.tresca@unipv.it