Performance Comparison of Jumping Gene Adaptations of the Elitist Non‐dominated Sorting Genetic Algorithm
Shivom Sharma, Seyed Reza Nabavi, Gade Pandu Rangaiah · 2013
Industrial problems are complex in nature, and often have multiple performance criteria. The elitist nondominated sorting genetic algorithm (NSGA-II) has been used to optimize many process design and operation problems for two or more objectives. In order to improve the performance of this algorithm, the jumping-gene concept from natural genetics has been incorporated in NSGA-II. Several jumping-gene adaptations have been proposed and used to solve mathematical and application problems in different studies. In this chapter, four jumping-gene adaptations are selected and comprehensively evaluated on a number of two-objective unconstrained and constrained test functions. Three quality metrics, namely, generational distance, spread and inverse generational distance are employed to evaluate the distribution and convergence of the obtained Pareto-optimal solutions at the final generation and also at selected intermediate generations. Additionally, a search termination criterion based on the improvement in the Pareto-optimal front, has been described and used to check convergence of NGSA-II with the selected jumping gene adaptations.