Setting parameters in the Genetic algorithm
The default parameters in GA should therefore be adequate for most problems. However, if the user needs to explore different methods, the following parameters may be set for GA. These are only available in the command input file.
GA parameters and default values:
Item | Parameter | Default value | Type | Remark |
popsize generation selection Tourn Size Elitism NumElites Encoding variable Numbits variable Binary crossover type Binary crossover probability Real crossover type Real crossover probability BLX alpha param | Population size Number of generations Selection operator: Tourn=1, Roullette=2, SUS=3 Tournament size for tournament selection operator Switch elitism for single objective GA Number of elites passed to next generation Type of encoding for a variable: Binary=1, Real=2 Number of bits assigned to a binary variable Type of binary crossover: Single point=1, Uniform=2 Binary crossover probability Type of real crossover: SBX=1, BLx=2 Real crossover probability Value of α for BLX operator | 30/100 100/250 1 2 1 2 2 15 1 1.00 1 1.00 0.5 | Integer Integer Integer Integer Integer Integer Integer Integer Integer Real Integer Real Real | 1 1 2 2 2 |
Real crossover distribution index | Distribution index for SBX crossover operator | 10.0 | Real | |
Binary mutation probability | Mutation probability for binary mutation | 1/number of binary digits | Real | |
Real mutation probability | Mutation probability in real- space | 10.0 | Real | |
Real mutation distribution index | Distribution index for mutation operator | 10.0 | Real | |
Restart interval | Frequency of writing restart file for direct GA For multi- objective problems, this parameter governs the frequency of writing TradeOff files | 10 | Integer | |
seed | Random seed | Long |