COMPARISON OF SOME PARALLEL GENETIC ALGORITHMS WITH POPULATION REINITIALIZATION
被引:0
|
作者:
Sekaj, I.
论文数: 0引用数: 0
h-index: 0
机构:
Slovak Tech Univ Bratislava, Fac Elect Engn & Informat Technol, Inst Control & Ind Informat, Bratislava 81219, SlovakiaSlovak Tech Univ Bratislava, Fac Elect Engn & Informat Technol, Inst Control & Ind Informat, Bratislava 81219, Slovakia
Sekaj, I.
[1
]
Oravec, M.
论文数: 0引用数: 0
h-index: 0
机构:
Slovak Tech Univ Bratislava, Fac Elect Engn & Informat Technol, Inst Control & Ind Informat, Bratislava 81219, SlovakiaSlovak Tech Univ Bratislava, Fac Elect Engn & Informat Technol, Inst Control & Ind Informat, Bratislava 81219, Slovakia
Oravec, M.
[1
]
机构:
[1] Slovak Tech Univ Bratislava, Fac Elect Engn & Informat Technol, Inst Control & Ind Informat, Bratislava 81219, Slovakia
In solving many practical search/optimization problems using evolutionary algorithms it is often difficult to avoid the premature convergence in search for the global optimum. From that reason parallel evolutionary algorithms (PEA) and Parallel Genetic Algorithms (PGA) can be used more effectively. In this paper some selected fine- and coarse-grained PGA architectures are analyzed and experimentally compared. Also the influence of population reinitialization on the parallel genetic algorithm performance is analyzed The results are demonstrated on the minimization of selected test functions.