Research
Education
Master of Science in Control Engineering
Advanced training in nonlinear systems, feedback control, system identification, optimization, and model-based engineering for complex dynamic systems. Awarded the DiSCo Lazio Scholarship for Academic Merit.
Bachelor of Science in Electrical Engineering – Control Systems
Core training in electrical engineering — circuits, signals and systems, electronics, and electromechanics — with a focused concentration in control systems and dynamic system modeling. This programme laid the mathematical and analytical foundation for graduate specialisation in control engineering at Sapienza University of Rome.
Academic Projects and Research
MSc Thesis: Nonlinearity Analysis in Reset Control Systems
Reset control systems — such as those based on the Clegg Integrator — improve transient performance by resetting the controller state to zero when certain conditions are met. However, this reset action introduces nonlinearity into an otherwise linear closed-loop system, causing standard frequency-domain analysis tools to produce unreliable results. The core challenge is that no unified, practical method existed to quantify how much nonlinearity a given reset controller introduces under real operating conditions.
This thesis proposed the coherence function as a quantitative indicator of nonlinearity in closed-loop reset systems — a criterion that measures the degree to which the system output remains linearly related to its input. A tuning approach was developed for Clegg Integrator-based controllers to reduce nonlinear effects, making frequency-domain analysis more reliable in practice. The coherence-based criterion was implemented and validated in MATLAB/Simulink, then compared against two established methods — the Describing Function and HOSIDF analyses — confirming coherence as a practical, unified indicator.
Methods and tools: MATLAB/Simulink · Frequency-domain analysis · Reset control · Coherence function · HOSIDF · Describing Function
Electrodynamic Shaker System Control
The core challenge in this project is the dynamic coupling between the electrodynamic shaker's actuator and the vehicle's mechanical response — the two subsystems influence each other, making the combined model significantly more complex than either in isolation. An LQR controller with a state observer was designed to suppress pitch vibration and reject disturbances, with performance evaluated not only at the nominal operating point but across a range of parameter variations — reflecting the realistic engineering condition where exact system parameters are never known precisely.
Methods and tools: State-space modeling · LQR · Observer design · Parametric uncertainty · MATLAB
Quadrotor Control via LQR
A quadrotor has six coupled degrees of freedom — position and orientation in three axes each — and is inherently unstable at hover without active control. The first step was linearizing the full nonlinear 6-DOF dynamics around the hover equilibrium, followed by a controllability analysis to confirm the linearized model was stabilizable before designing any feedback. LQR feedback gains were then computed and trajectory-tracking performance was evaluated in simulation.
Methods and tools: Linearization · Optimal control · Controllability analysis · MATLAB/Simulink
Digital Control of a Second-Order Plant
Designed continuous-to-discrete controllers and observers, and evaluated the effect of sampling on closed-loop stability margins and transient behavior.
Methods and tools: Discretization · Deadbeat control · Observer design · MATLAB
Artifact Image Classification
Compared Bag-of-Features and transfer-learning pipelines for multi-class classification of archaeological artifacts, evaluating model behavior across different visual feature representations.
Methods and tools: ResNet50 · GoogLeNet · Bag of Features · MATLAB
For research collaboration, contact [email protected]