COSTA JUNIOR, A. G.; http://lattes.cnpq.br/0585826541984052; COSTA JÚNIOR, Ademar Gonçalves da.
Resumo:
The Linear Variable Differential Transformer (LVDT) is a displacement sensor, used for
linear displacement measurements. The displacement of the LVDT core varies the mutual
inductance between the primary and secondaries coils, consequently, there is a voltage in
the secondary coils, proportional to this displacement. The conditioning of this signal
is used to compatibility of the data acquisition system. A commercial LVDT is sold
together with a signal conditioner, being this proper system manufacturer. The system
is calibrated by manufacturer and so, used by user.
Many techniques have been proposed for signal conditioning LVDT. In this dissertation,
presents for extracting of the displacement signal from the LVDT sensor, the use a
model based system identification technique. The LVDT mathematical model is obtained
through application Voltage Kirchoff’s law, to the equivalent electrical circuit. To parameter
estimation of the LVDT is used the time-continuous model. Carry out the analysis
of parametric sensibility in the time-continuous model from LVDT. The estimation algorithms
used in this dissertation are least squares criterion and its recursive version.
Through the development of two methods, that utilize the parameters of LVDT‘s model,
as a displacement function, estimate the displacement signal from its core. The technique
proposed is valued at numerical simulation. Finally, the technique is valued at through
experiments, where carry out a similar analysis to the simulation results, which shows the
viability the concept this implementation.