CELLULAR AUTOMATA ARRAY
| Collection | International Encyclopedia of Systems and Cybernetics |
|---|---|
| Year | 2004 |
| Vol. (num.) | 2(1) |
| ID | ◀ 388 ▶ |
| Object type | Methodology or model |
- “… an interconnected collection of similar subunits … (each of which is) usually not to be more complex than a finite automaton” (p.45).
A. LINDENMAYER and K. CULIK II define this model and state some shortcomings of former models of the kind: “In the case of the McCULLOCH-PITTS nerve net models, the problem is that the size of the array remains constant; furthermore, that the entire array behaves as a single finite automaton. In the case of von NEUMANN's self-reproducing automata, the array is laid out on a square grid and is allowed to grow only at the margins (i.e. the active part of the array). While these structures are as powerful as TURING machines, they are not suited to simulate in a realistic manner growing cells or multicellular organisms” (Ibid).
To remedy these limitations, the authors proposed a modeling method whose principle is what they call parallel rewriting of graphs.
The formalism they propose seems closely related to the specific problem of internal growth of networks, as for example in embriology and more generally in morphogenetic fields. The subject is also closely related to the “Physics of vegetal spirals” and FIBONACCI series (see St. DOUADY and Y. COUDER — 1993).