Development of a basic code for multi-objective genetic algorithms

U ovom radu su obrađene metode optimiranja, s naglaskom na genetske algoritme i višeciljnu optimizaciju. Napisana su dva programa za višeciljnu optimizaciju i detaljno razložena po dijelovima. Prvi program konstruira Pareto frontu koristeći jedinstvenu funkciju cilja i variranje težinskih koeficijenata, a kao metoda traženja optimuma jedinstvene funkcije cilja koristi se jednociljni genetski algoritam. Drugi program koristi MOGA-u. Programi služe za demonstraciju metoda i nisu zamišljeni za praktičnu primjenu. Za testiranje ispravnosti programa koristi se Binh i Korn testna funkcija.This thesis is about optimization, with focus on genetic algorithms and multi-objective optimization. Two programs have been written for multi-objective optimization and are broken down in detail part by part. The first program constructs the Pareto front by using consolidated single excellence criterion and variation of weight factors. Single-objective genetic algorithm is used to find the optima of the single excellence function. The other program uses MOGA. The programs are made to demonstrate the methods and were not made for practical use. To test the functionality of the programs, Binh and Korn test function is used