An adaptive annealing genetic algorithm for the job-shop planning and scheduling problem

Abstract

The genetic algorithm, the simulated annealing algorithm and the optimum individual protecting algorithm are based on the order of nature, and there exist some application limitations on global astringency, population precocity and convergence rapidity. An adaptive annealing genetic algorithm is proposed to deal with the job-shop planning and scheduling problem for the single-piece, small-batch, custom production mode. In the AAGA, the adaptive mutation probability is included to improve upon the convergence rapidity of the genetic algorithm, and to avoid local optimization, the Boltzmann probability selection mechanism from the simulated annealing algorithm, which solves the population precocity and the local convergence problems, is applied to select the crossover parents. Finally, the AAGA-based job-shop planning and scheduling problem is discussed, and the computing results of AAGA and GA are depicted and compared.