In order to solve the problem of insufficient modeling of the relationship between the batch division decision and the production cost in the existing flexible job shop batch scheduling, a multi-objective evolutionary algorithm NSGA2 _ SA is proposed, which integrates the NSGA2 population evolution mechanism and the simulated annealing algorithm. a three-segment coding scheme is designed to represent batch partition, machine selection and process sequencing; an adaptive mutation operator based on dynamic processing information is used to improve the search efficiency; a simulated annealing algorithm is introduced to enhance the local search ability. The experimental results of small, medium and large scale examples show that the proposed algorithm has high efficiency and stability.