- author = "William Comisky and Jessen Yu and John R. Koza",
- title = "Automatic Synthesis of a Wire Antenna Using Genetic Programming",
- booktitle = "Late Breaking Papers at the 2000 Genetic and Evolutionary Computation Conference",
- year = "2000",
- editor = "Darrell Whitley",
- pages = "179--186",
- address = "Las Vegas, Nevada, USA",
- month = "8 " # jul,
- keywords = "genetic algorithms, genetic programming, voltage standing wave ratio, VSWR, Gain, Yagi-Uda wire antenna, Altshuler-Linden, Logo programming language, L-systems, wire printing, parallel computer",
-
URL = "
http://www.genetic-programming.com/jkpdf/gecco2000antenna.pdf",
- size = "9 pages",
- abstract = "This paper demonstrates the use of genetic programming to automatically synthesize the design of a wire antenna for an illustrative problem that has been previously solved by both conventional antenna design techniques and the genetic algorithm operating on fixed-length character strings. When the genetic algorithm was used, the human user prespecified many characteristics of the size and shape of the solution. The run of genetic programming also produced a satisfactory result for the illustrative problem. However, it did not require the human user to prespecify the size and shape of the solution. Functions from the Logo programming language and Lindenmayer systems enable genetic programming to draw the antenna. The solution evolved by genetic programming possesses the essential characteristics of the Yagi-Uda type of antenna. The rediscovery by genetic programming of the essential characteristics of the Yagi-Uda antenna is an instance where genetic programming has produced a result that is competitive with a result produced by creative and inventive humans.",
-
notes = "TURN-RIGHT, DRAW, REPEAT, LANDMARK, TRANSLATE-RIGHT,
PROGN2, PROGN3, PROGN4. WIRE, NOWIRE, random real, and
random integer, END.
Numerical Electromagnetics Code antenna simulator, NEC, fortran code.
Ie three D 3D print conductive wire in NEC simulation.
Linux Beowulf-style parallel cluster computer system. Distributed imigration-emigration deme parallel population. Two percent mutation.
22 hours = 2.772e+16 computer cycles = 28 peta-cycles of computer time
Part of \cite{whitley:2000:GECCOlb}",
