Impedance spectroscopy lives with a tension between speed and accuracy. My work addresses both at once, by designing the excitation signal and the estimation algorithm together.
Equivalent-circuit modeling and Distribution of Relaxation Times (DRT) to extract physical meaning from spectra.
Kalman filter variants and particle filters for tracking and parameter extraction under noise.
Meta-heuristics and gradient-based methods, model-aware feature selection (Fast A*-mRMR), and machine learning from shallow models to deep networks (CNN–Transformer) for diagnostics and state estimation.
Topics under active investigation with my doctoral researchers and students.
Deep-learning state-of-charge estimation (CNN–Transformer hybrids), model-aware feature selection, and embedded ML for battery diagnostics.
Microcontroller-based (STM32) impedance spectroscopy with resource-optimized real-time DSP, low-power design, and on-board deployment for portable and field systems.
From measurement to action: impedance-informed battery management, control strategies, and knowledge-structured (ontology-based) measurement systems.
More accurate, faster state-of-health and state-of-charge estimation for electric mobility and stationary energy storage.
Advances in bioimpedance spectroscopy supporting non-invasive diagnostics and wearable monitoring.
Continuous monitoring of power cables, bus systems, and bridges — reducing downtime and enabling predictive maintenance of critical infrastructure.