1991 Fiscal Year Final Research Report Summary
Study of dynamics of complex systems from the view point of information processing
| Project/Area Number |
02302022
|
|---|
| Research Category |
Grant-in-Aid for Co-operative Research (A)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
物性一般（含極低温･固体物性に対する理論）
|
|---|
| Research Institution | Kyoto Univeristy |
|---|
Principal Investigator |
IKEDA K. Kyoto University, Research Inst. for Theoret. Phys., Prof., 基礎物理学研究所, 教授 (40151287)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MATSUMOTO K. Hokkaido Univ. Faculty of PHARMACEUTICAL Sciences, Instructor, 薬学部, 教務職員 (80183953)
SANO M. Tohoku Univ, Research Inst. of Electric Engineering, Associate Prof., 電気通信研究所, 助教授 (40150263)
KANEKO K. Univ. of Tokyo, College of General Arts and Sciencess, Associate Prof., 教養学部, 助教授 (30177513)
TSUDA I. Kyushu Inst. of, Technology, Faculty of Computer Sciences, Asoociate Prof., 情報工学部, 助教授 (10207384)
|
|---|
| Project Period (FY) |
1990 – 1991
|
| Keywords | Chaos / Chaotic itinerancy / Information theory / Turbulence / Neural network / Liquid crystal / Coupled map lattice / Slime mold |
|---|
| Research Abstract |
The aim of the project is to elucidate complex dynamical behavior in chaotic systems with infinite number of degrees of freedom from the view point of information processing underlying the system's dynamics
The first aim is to establish a method based upon the information theory with which the spatio-temporal structure of information flow through complex dynamical medium may be read out. We developed a method of information theoretical analysis by introducing mutual information (MI), information flow rate, multi-port MI etc, and applied it to biological system (slime mold), and physical systems (laser-optical turbulence, and fluid turbulence). We confirmed that the information theoretical method is very powerful in elucidating the dynamical connectivity among a huge number of elements each of which contribute in a complicated manner to the motion exhibited by the whole system.
The second aim is to make clear the significance and mechanism of complex dynamics from the view point of information processing. In particular we forcus our attention to chaotic itinerancy (CI) which is a chaos-induced itinerant motion among the ruins of local attractors and is commonly observed in infinite dimensional systems. A most interesting result is that the itinerancy is a highly organized motion very different from the noise-induced transition processes. The organized aspects of chaos yielding itinerancy are extensively studied in a nonlinear optical system, a neural network model of brain, and coupled map lattices. A possibility that CI may improve the functions of neural network is pointed out. Experimentally, the life cycle of a target pattern self-organized on a 2D-lattice of thermal convection pattern is investigated with a liquid crystal.
|
