[Colloq] Colloquium talk Wednesday 2/13 2:00pm 149CN

[Rachel Bates]

*****************    CCS COLLOQUIUM    **********************

Wednesday, 13 February, 2002
2:00-3:00 p.m.
149 CN

ADAPTATIONAL POSSIBILITIES OF COMPLEX SYSTEMS WITH UNCERTAINTY
Mikhail B. Ignatyev
Head of Computer Science Department
St-Petersburg State University of Aerospace Instrumentation


Abstract

Systems must work in complex environment and must have good adaptational
capabilities.  Low organized environment demands systems with high
adaptational capabilities, highly organized environments permits to use the
systems with low adaptational capabilities, but the creation of highly
organized environment is very expensive. In our lecture we investigate the
adaptational capabilities of different systems by means of
linguo-combinatorial simulation method.  We prove, that adaptational
possibilities characterize the number of arbitrary coefficients S in
equivalent equations of system
                                m+1
                        S = C          n > m                   ( 1 )
                                n
where n is the number of system variables for example - the number of
degrees of freedom (a man's body has more one hundred degrees of freedom),
and m is the number of restrictions.  The number of arbitrary coefficients
is the measure of uncertainty.  To have the best adaptational capabilities
we must be in the adaptational maximum zone.  The strategy of system's
control must strive to hold itself in the adaptational maximum zone.  We
investigate different approaches for decision of this problem by means of
system learning, collection of joint of systems and other. We create an
initial set of equivalent equations based on natural language description
developed by selecting key words. Equation (1) is the main law of
cybernetics. In the lecture we will consider different applications such as
modeling city development, global model of organism for decision-making
support, robotic control system, game control, and virtual worlds. Computer
simulation of different system types illustrates different effects of
adaptation.


About the Speaker:
Professor Mikhail B. Ignatyev, Doctor of technical science, Honorable
scientist of Russia, Laureate of the State Prize of USSR, is Head of
Computer
Science Department of St-Petersburg State University of Aerospace
Instrumentation, Director of Independent Institute of Sustainable
Development, and President of Virtual Reality Society of Russia.  He is a
member of Natural Science Academy, Metrological Academy, Non-Linear Science
Academy, and a senior member of Society for Computer Simulation
International.  He is a pioneer in the field of robotics and supercomputing
and was a vice chief designer of Robotics in USSR from 1972 to 1991.  He
published more than 450 scientific articles, inventions and books. Some of
the titles are "Holonomical automatic systems" (in Russian), "Algorithms for
Robot Manipulator Control" (Russian and English), "Computing process
control" (in Russian), "Underwater robots" (in Russian), "Simulation of
machine systems" (in Russian), "Models and systems for complex experimental
analysis" (in Russian), and "Active method of reliability guarantee of
algorithms and programs" (in Russian).  His research interests are in the
complex systems simulation, adaptation and virtual world architecture.
If you would like to meet with the speaker, please send a note to
lorenz@ccs.neu.edu.

Host:
David H. Lorenz