Naslov
Performance Analysis-based GA Parameter Selection and Increase of μ GA Accuracy by Gradual Contraction of Solution Space

Autori
Dužanec, Darko ; Kovačić, Zdenko

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of the 2009 IEEE International Conference on Industrial Technology (ICIT) / - Melbourne, 2009

Skup
The 2009 IEEE International Conference on Industrial Technology (ICIT)

Mjesto i datum
Gippsland, Australija, 10.02.2009. - 13.02.2009

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Microgenetic algorithms; inverse kinematics; robot control

Sažetak
Although methods for design of genetic algorithms (GA) are well established, general expressions for determination of optimal GA parameters are still missing. There is also a problem of possible inaccuracy of a found solution. This paper describes a GA performance analysis for a selected vector-based optimization problem that has led to useful GA parameter selection criteria. The paper also describes a new method for increasing the precision of a complementary micro genetic algorithm (μ GA) by enforcing gradual contraction of the space of candidate solutions during optimization. The enhanced μ GA has been tested on the model of a 13-DOF tentacle robot, and the performance analysis showed significant improvement of accuracy without affecting the duration of the algorithm.

Izvorni jezik
Engleski

Znanstvena područja
Temeljne tehničke znanosti

POVEZANOST RADA