Providing a genetic algorithm optimized based on regression equations for solving flexible jobshop scheduling problem (With objective function minimization for makespan)

摘要

This study includes FJSP with objective function minimization for makespan. In this study a Meta heuristic algorithm based on genetic algorithm method for solving is presented. So that the result from this algorithm is based on targeted and effective combination of number of produced solutions obtained through operators of genetic algorithm. Also a multivariate equation is introduced based on existing relationship (between total solutions provider of new generation about each mutation operator, integration and cross over using regression equations). Then a linear regression equation with the lowest total square error for the relationship between number of population and number of generations produced by three operators mentioned to produce the next generation is obtained. These results in order to evaluate the effectiveness and efficiency of proposed algorithm are compared with optimal solutions obtained through the Lingo software and other similar research results. Results of experiments and computational analysis show that the proposed algorithm in this study has ability to achieve close to optimal points for different issues in different sizes.