Personal pages
Contact information
email Flavio Stellino
Via Opera Pia 11a,
16145, Genoa, Italy
tel: +39-010-353-2276
fax: +39-010-353-2290

Flavio STELLINO, Ph.D. student

last modified July 22, 2016 - 14:01 CEST

Complex Nonlinear Dynamical System with emphasis on:

  • Identification and Compensation of Dynamical Hysteretic Systems. In this work I defined an analytical model of dynamical hysteresis, identified from experimental data, collected on a Piezoelectric Actuator (PEA) [3,5]. Nanopositioning devices are based on PEAs, which exhibit Stress-Strain characteristic subject to hysteresis and creep, nonlinear phenomena which compromise positioning accuracy. In order to compensate for these phenomena I employed an open loop control [1,4], based on a feedforward controller, called compensator, defined by an analytical inversion of the proposed model. Furthermore, this model has been compared with other models of hysteresis and creep [2,6].
  • Neuron Models: I analysed different deterministic neuron models through bifurcation theory, to reproduce qualitatively different neural responses. Furthermore I considered stochastic neuron models, by solving Langevin equations or simulating Markov Models, in order to reproduce a more realistic behaviour of the neural response, subject to channel noise and synaptic noise.
Neural Coding: I employed neuron models to study neural communication codes on two different aspects:
  • Ability of neurons to transmit information. I applied Information Theory to estimate Shannon Entropy and Mutual Information, that quantify in absolute terms the transmitted information of neural spike trains in response to a synaptic stimulus. This approach has been employed also when the neuron works in bursting conditions.
  • Study of neural stimuli encoding. Action Potential is the stereotyped neural response which codes synaptic input stimulus. The main problem of Neural Coding concerns the identification of stimulus parameters (features) which modulate neural response, transmitting information. In my work I studied phase codes, evaluating correlation between the phase of the input stimuli (computed by Hilbert Transform) and the neural response. This work is in progress in collaboration with Alberto Mazzoni of the "Scuola Superiore sant'Anna of Pisa" and it will be soon submitted to a journal for publication.

This site conforms to the following standards: