COSTA, M. M.; http://lattes.cnpq.br/0739428363982644; COSTA, Maxwell Moura.
Resumen:
The measurement of wind speed is of great importance in various fields of knowledge, its
value combined with low measurement uncertainty can, for exemple, it is used to determine
the best locations for installing wind turbines. Thus, this work presents a new methodology
wind speed measurement based on ultrasonic transducers, statistically estimating its outcome.
The principle of operation of ultrasonic transducers is based on measuring the Time-of-Flight
(ToF), which can be estimated using various techniques, such as: Threshold Detection (TH),
Phase Difference (PD), Data Fusion (MLE) and Cross-Correlation (CC). Although there are
different techniques, research has been developed for determining the ToF with lower
uncertainty measurement. In this Thesis, the ToF was estimated using Discrete Fourier
Transform (DFT), which from its symmetry properties, periodicity and analysis of signals in
the frequency domain, allows for an estimate of the difference between the phases of the
ultrasonic signals transmitted and received, consecutively obtained the ToF. In order to
evaluate the proposed technique, we designed a mathematical model for an ultrasonic system,
based on the principles of system identification theory. Thus, a set of experimental data
collected from the ultrasonic frequency response system was used for the preparation of a new
mathematical model, using both the method of Levi to identify the system parameters,
resulting in a model with an approach index of approximately 90%. In order to evaluate the
quality of the ToF estimation, an analysis of the propagation phase difference of the
uncertainties in the presence of multiplicative noise and additive was performed, obtaining an
expression that characterizes them. It was observed that only the additive noise has an
influence on phase measurement uncertainty. Thus, there was a study to assess the influence
of additive noise in the signal phase of uncertainty, which culminated in the development of a
new expression which assesses the uncertainty phase and, therefore, the spread of
measurement uncertainty. The results of computational simulation and experimental show that
the ToF estimation technique based on Discrete Fourier Transform provides a smaller
uncertainty measurement when compared with conventional techniques, consequently the
estimates of ToF were consistent with the classical estimation techniques confirming the
adequacy of the new technique for use in the estimation of ToF and therefore the wind speed
estimation.