Probe Characterization Using Stochastic Search Algorithms and Radial Basis Functions Neural Networks

Within this paper, the auto-compensated electrostatic induction probe is characterized by generating its Point Spread Function (PSF) and its inverse. New approaches based on the stochastic Search Algorithms (SSA) are proposed to obtain several Point Spread Functions and their inverses. Radial Basis...

Full description

Saved in:
Bibliographic Details
Published in2022 19th International Multi-Conference on Systems, Signals & Devices (SSD) pp. 1044 - 1050
Main Authors Melahi, Ahmed, Bendahmane, Boukhalfa, Neagoe, Marian Bogdan, Dascalescu, Lucian
Format Conference Proceeding
LanguageEnglish
Published IEEE 06.05.2022
Subjects
Costs
Electric potential
Electrostatic induction
Electrostatic Induction Probe
Electrostatic measurements
Performance evaluation
Point Spread Function
Prediction algorithms
Radial basis function networks
Radial Basis Functions Neural Networks
Stochastic Search Algorithms
Surface Potential
Online AccessGet full text

Cover

More Information
Summary:Within this paper, the auto-compensated electrostatic induction probe is characterized by generating its Point Spread Function (PSF) and its inverse. New approaches based on the stochastic Search Algorithms (SSA) are proposed to obtain several Point Spread Functions and their inverses. Radial Basis Functions Neural Networks are used to approximate the original measures in order to have two-dimensional array representation of the potential on the surface of the material. These new approaches are tested, validated, and compared to each other in term of rapid convergence and best cost obtained. The obtained PSFs and inverse PSFs give approximations within a good accuracy in predicting the measure or the real distribution of the potential on the surface of the material.
ISSN:2474-0446
DOI:10.1109/SSD54932.2022.9955944