Study of Operator's adaptability and scale-up study for RAGA

The studies have shown that a class of recombination operators is more suitable to tackle certain problems than others. It is observed that the multi-parent recombination operator with polynomial distribution (MPX) is exploitative and the multi-parent recombination operator with lognormal distribution (MLX) is explorative, in nature. Use of productive operators is necessary for a genetic algorithm to uncover new fitter points in the search space to improve its overall performance. Real-coded self-Adaptive GA (RAGA) uses two multi-parent recombination operators (MPX and MLX). The use of particular operator to generate offspring during evolution process depends on its ability to produce good offspring. This paper presents the effect of operator's adaptability in solving test problems. Also scale-up study analyses the performance of RAGA with increasing number of control variables.