학술논문

MEASUREMENT THEORY AND SIGNAL THEORY PART I - Investigations on the Foundations
Document Type
Article
Source
International Journal of General Systems; December 1984, Vol. 10 Issue: 1 p15-26, 12p
Subject
Language
ISSN
03081079; 15635104
Abstract
The concept of a sensor with finite resolution is chosen as the cornerstone of a physical measurement theory. It is concluded that the concept of a band-limited channel is less appropriate for this purpose because it is defined with help of specific mathematical operators on functions that are defined on an infinite interval.It is shown that measurement in continuum physics requires transfer, transformation and transport of energy to, by, and through a sensor: a non-equilibrium process. As theory requires equilibrium for any measurement, a paradox arises. It is indicated that experimental physics avoids this paradox with help of a hierarchy of space and time scales. The definitions and propositions are formulated in a similar way in order to avoid this equilibrium/non-equilibrium paradox.The equivalence of the various forms of energy, as stated by the first law of thermodynamics and verified by experimental physics, renders interpretation of physical measurement relatively easy.Signals are defined to be products of measurement. They are carriers of a finite amount of information. As the class of continuous functions generate an infinite amount of information, discrete functions are selected to represent sampled and digitized signals. If the sampling interval is of the order of the largest characteristic time scale of the total system, then little information is lost by digitization. The number of significant digits is related to the resolution of the sensor and of the other parts of the measuring instrument. A measuring instrument is shown to be composed of a chain or a set of chains of sensors.