Modified adaptive genetic algorithms for solving job-shop scheduling problems

Job-shop scheduling problem (JSP) is one of the most difficult combinatorial optimization problems. It is widely applied to productive management of enterprise. This paper proposed improved adaptive genetic algorithms for solving job-shop scheduling problems according to the idea that the best individual on current generation should be kept to next generation, but the best individual should be crossed and mutated by some probability. The software package for these modified adaptive genetic algorithms is programmed and applied to solving job-shop scheduling problems. These modified methods increase the convergence rate. Especially, the crossover probability and mutation probability are given automatically in the search process.