The paper aims to present a brain storm optimization(BSO) algorithm integrating the population diversity and discussion mechanism(PD-DMBSO) to solve the flexible job shop scheduling problem(FJSP). Firstly, a math optimization model for FJSP is built. Secondly, the flowchart of the PD-DMBSO is proposed. Then, the key parameters of the PD-DMBSO are discussed. Considering the characters of the FJSP, the searching operators are designed. An extended operation encoding method is proposed. The K–means clustering algorithm is employed to cluster the individuals. A perturbation operator and a combining operator are designed to generate new individuals. Finally, a group of experiments are conducted to compare the four algorithms. The statistics analysis of the experiment results shows that the PB-DMBSO performs best among the four algorithms because it can effectively avoid the premature convergence and ensure to explore more solution space for discrete production scheduling problem.