An adaptive genetic algorithm with dominated genes for distributed scheduling problems

This paper proposes an adaptive genetic algorithm for distributed scheduling problems in multi-factory and multi-product environment. Distributed production strategy enables factories to be more focused on their core product types, to achieve better quality, to reduce production cost, and to reduce management risk. However, when comparing with single-factory production, scheduling problems involved in multi-factory one are more complicated, since different jobs distributed to different factories will have different production scheduling, consequently affect the performance of the supply chain. Distributed scheduling problems deal with the assignment of jobs to suitable factories and determine their production scheduling accordingly. In this paper, a new crossover mechanism named dominated gene crossover will be introduced to enhance the performance of genetic search, and eliminate the problem of determining optimal crossover rate. A number of experiments have been carried out. For the comparison purpose, five multi-factory models have been solved by different well known optimization approaches. The results indicate that significant improvement could be obtained by the proposed algorithm.