A multi-objective integrated genetic algorithm(MIGA) was proposed to solve both of the assignment of machines to corresponding operations and the scheduling of operations on assigned machines in flexible job scheduling problem at the same time.MIGA used random weights to convert the multi-objective problem to a single-objective problem. An operation-extended coding method was presented.The elitist strategy and the niche technology were integrated into the roulette selection operation to speed up the convergence and to improve the diversity of the population respectively.A new scheduling-batch oriented active decoding method was designed.Finally,some benthmark problems were experimented.The results prove that MIGA can solve mass variety and small batch scheduling problem in the flexible job shop effectively and efficiently. 

提出一种多目标混合遗传算法(MIGA)，采用集成法同时解决柔性作业车间调度的两个子问题：机器分配问题和工序调度问题。MIGA在标准遗传算法的基础上采用随机权重法解决多目标问题，引入精英保留策略加速算法的收敛，集成小生境技术提高种群的多样性，基于扩展工序编码，按Makespan和安装准备成本最优对调度批分别解码。最后，用标准算例进行了算法验证，证明MIGA可以有效解决柔性作业车间多品种小批量调度问题。

柔性作业车间；多品种小批量；调度；多目标混合遗传算法, 调度批