Abstract: In order to solve the distributed heterogeneous flexible flowshop rescheduling problem (DHFFRP), considering machine breakdown, an integer planning model with the objective of minimizing the maximum completion time is established, and a hybrid artificial bee colony algorithm (HABCA), which incorporates genetic algorithms, variable neighborhood search strategies, and the idea of simulated annealing, is proposed for problem-solving. In order to cope with the dynamically changing process states, the above dynamic problem is transformed into a series of static problems with phases by distinguishing different states of the process according to the failure time and the faulty machine. In HABCA, a three-layer coding method and an insertion-based greedy decoding mechanism are designed; in the employed bees search process, crossover and mutation strategies are introduced to optimize the nectar sources in the population; in the onlooker bees search process, a local search operator based on the minimum completion time rule is used; and in the scout bees search process, four kinds of variable neighborhood search operators and simulated annealing operations are applied for the evolution of the solution. The performance improvement between initial scheduling and rescheduling is verified through a large number of example tests; the proposed algorithm is compared with the classical artificial bee colony algorithm and existing algorithms, and the results show the effectiveness and superiority of HABCA.
Keywords: distributed heterogeneous flexible flowshop; machine breakdown; rescheduling; hybrid artificial bee colony algorithm(HABCA); simulated annealing operation
合肥工业大学学报自科版