Improved Hormone Algorithm for Solving the Permutation Flow Shop Scheduling Problem

Based on the large objective function value imbalance problem of selection, crossover and mutation in hormone-regulation based genetic algorithm, an adaptive genetic algorithm based on improved hormone concentration calculation method (referred to as IHCCM-IAGA) is proposed. IHCCM-IAGA adopts the coding method based on the arrangement of work pieces, and uses the reverse learning method to initialize the population, which improves the quality of the initial solution. Aiming at the problems of high redundancy and low efficiency in the two-points crossover (TPX) operator, an improved TPX (improved two-points crossover, ITPX) is proposed, and the introduction of excellent gene pool and immune factors realizes two crossover methods and monitors the entire evolution process, avoiding the loss of high-quality chromosomes. A variety of perturbations are designed to maintain a rich diversity structure and related local search algorithms are combined into a mutation operator. A population annihilation operator is established and an annihilation factor is set to guide the local search in the mutation operator. IHCCM-IAGA is applied to the permutation flow shop scheduling problem, and various tests of the standard calculation example of the problem are performed. The results show that IHCCM-IAGA is effective.