A Direction-Based Exponential Mutation Operator for Real-Coded Genetic Algorithm
Amit Kumar Das, Dilip Kumar Pratihar · 2018
In genetic algorithm (GA), mutation is one of the most important operators responsible for maintaining diversity in the population. For a real-coded genetic algorithm (RCGA), this mutation operator is applied variable-wise. In this study, a new direction-based exponential mutation operator (DEM) has been proposed for an RCGA. This newly developed operator is influenced by the directional information of the design variables. The locations of mutated solutions are made biased to the said information of the problem with the higher probability value. The performance of the proposed operator with an RCGA has been tested on twenty classical benchmark optimization functions and these results are compared with that of an RCGA with polynomial mutation operator. The said comparison clearly shows the superiority of the proposed operator.