|
R. Belu, A. Paun, A. R. Belu
Neural Networks in Measurement, Instrumentation and Control
Abstract.
Artificial intelligence techniques are now attractive tools used to enhance and
improve the efficiency, the capability, and the features of the instrumentation
in application areas related to measurements, system identification, and
control. These techniques exploit the computational capabilities of modern
computing systems to manipulate the sampled input signals and extract desired
measurements. The aim of this paper is to introduce the fundamental concepts of
artificial neural networks and to present a review of those applications related
to measurement, instrumentation and control.
Artificial neural networks have several advantages in practice, including
learning and adapting ability, parallel distributed computation, robustness,
etc. The use of artificial neural networks in instrumentation, measurement, and
control applications has been increasing dramatically in the last three decades.
This paper describes some of the neural network techniques providing the reader
with in-depth theoretical elements and references to allow them to explore the
applicability of artificial neural network technologies for their specific
applications. One of the main topics of this paper will refer to surveying the
neural signal understanding and classification methods, and the selected
implementation and architecture will be described with its performance in terms
of correct classification rates and robustness to noise. The chosen forms of
neural networks were tested using different types of input information. This
paper is the first in a series of three papers dealing with problems related
with applications of neural artificial networks in measurement, instrumentation
and control. |