Theoretical Models for Thermoregulation in Man
263
which systems analysis uses can be applied to quite complex biological
control problems, but both the complexity and the non-linear characteristics
make it virtually impossible to use the formal theory which systems analysis
has developed. It is, perhaps, even unwise to attempt to force simplified
biological control concepts into the constraints which apply to control
systems theory.
It is perhaps superfluous to point out the major difference between the
use of the approaches of systems analysis in engineering and in biology.
The ultimate aim in engineering is a complete and general mathematical
prediction of the behavior of a complex system consisting of components
with known characteristics and known interrelationships. In biological
research, we deal with a complex system of which the behavior can be
investigated, and we attempt to gain insight into characteristics of components and the nature of their interrelationships. The number of such anatomical and functional components is large, and their interrelationships
are not all known in a quantitative way or even a qualitative sense.
Embedded in the literature we can find many isolated observations on
the characteristics or relationships of a single component. In intuitive
evaluation of complex s\'stems, such observations are of small valueeven if they are recollected. The great contribution which the approach of
systems analysis can make is to provide a rigorous framework which allows
us to transfer such isolated observations with all its quantitative consequences
into a structure where it is combined with many more similar observations.
The completed mathematical structure then yields a predictive capability
for the whole system which is very enlightening and has recently been
recognized to be of considerable value in evaluation of applied problems.
Development of a Quantitative Model
A model of a complex system, such as thermoregulation, develops
gradually and is never complete: it should be a continuing process, combined
with a program of experimental work in the same field.
In our experience, after a series of small additions and refinements, it
usually becomes necessary to re-formulate the mathematical description as
insight improves or as new data become available. The initial definition is
conveniently expressed in the form of a block diagram. The simplest
division of the system is into a controlling system and a controlled system
as shown in Figure 1. Both blocks of Figure 1 can be subdivided into a
number of components. Obviously the final diagram will still represent a
very considerable simplification as compared with the real system. A model
must always be simpler than the system it represents if it is to fulfill a major
requirement of a model: it must be easier to understand and study than the
systems it represents. The development which such a model undergoes is
263
which systems analysis uses can be applied to quite complex biological
control problems, but both the complexity and the non-linear characteristics
make it virtually impossible to use the formal theory which systems analysis
has developed. It is, perhaps, even unwise to attempt to force simplified
biological control concepts into the constraints which apply to control
systems theory.
It is perhaps superfluous to point out the major difference between the
use of the approaches of systems analysis in engineering and in biology.
The ultimate aim in engineering is a complete and general mathematical
prediction of the behavior of a complex system consisting of components
with known characteristics and known interrelationships. In biological
research, we deal with a complex system of which the behavior can be
investigated, and we attempt to gain insight into characteristics of components and the nature of their interrelationships. The number of such anatomical and functional components is large, and their interrelationships
are not all known in a quantitative way or even a qualitative sense.
Embedded in the literature we can find many isolated observations on
the characteristics or relationships of a single component. In intuitive
evaluation of complex s\'stems, such observations are of small valueeven if they are recollected. The great contribution which the approach of
systems analysis can make is to provide a rigorous framework which allows
us to transfer such isolated observations with all its quantitative consequences
into a structure where it is combined with many more similar observations.
The completed mathematical structure then yields a predictive capability
for the whole system which is very enlightening and has recently been
recognized to be of considerable value in evaluation of applied problems.
Development of a Quantitative Model
A model of a complex system, such as thermoregulation, develops
gradually and is never complete: it should be a continuing process, combined
with a program of experimental work in the same field.
In our experience, after a series of small additions and refinements, it
usually becomes necessary to re-formulate the mathematical description as
insight improves or as new data become available. The initial definition is
conveniently expressed in the form of a block diagram. The simplest
division of the system is into a controlling system and a controlled system
as shown in Figure 1. Both blocks of Figure 1 can be subdivided into a
number of components. Obviously the final diagram will still represent a
very considerable simplification as compared with the real system. A model
must always be simpler than the system it represents if it is to fulfill a major
requirement of a model: it must be easier to understand and study than the
systems it represents. The development which such a model undergoes is