Abstract: To cope with the impact of multiple disturbances in production processes on actual scheduling processes, a multiple dynamic scheduling model for flexible job shops is established with urgent orders and machine breakdowns as disturbance factors and with the objectives of minimizing the makespan, order delay penalties and carbon emissions. A hybrid event- and cycle-based dynamic scheduling strategy is used to cope with emergencies, and an improved balanced optimizer algorithm is proposed to solve the model, which improves the initial population quality by adopting a hybrid population initialization strategy based on elite reverse learning. By using IPOX crossover, MPX crossover and mutation operations, the breadth and diversity of the algorithm is improved. An elite selection strategy based on Metropoils criteria is utilized to update the population and prevent it from falling into local optima. The searching ability of the algorithm is improved by double-layer variable neighborhood search. The effectiveness, stability and superiority of the algorithm are verified through a large number of extended numerical simulations.

Key words: hybrid equilibrium optimizer algorithm, multiple dynamic scheduling, variable neighbourhood search, elite reverse learning