Abstract: To efficiently analyze the performance of flexible flow shop with non-equivalent parallel Automatic guided vehicle (AGV),an open queuing network model with finite buffers based on stochastic process theory was established by taking the uncertainty of manufacturing system as well as the randomness of AGV transporting lot size fully into consideration.An improved state space decomposition method was proposed,and the state-space model was built to analyze the coupling effect between non-identical parallel AGVs and work units before and after them on this basis.The state transferring discipline was analyzed with Markov process theory.An iteration algorithm and the calculation method of system performance were proposed.The example results were compared with the statistical results of simulation model as well as the former researches to validate the accuracy of the improved state space decomposition method,and the sensitivity analysis for the system performance were also implemented.

Key words: automatic guided vehicle, flexible flow shop, queuing network, state space decomposition

摘要： 为有效分析具有非同等并行自动导引小车的柔性流水车间系统的性能,考虑制造系统的不确定性以及自动导引小车运输批量的随机性,应用随机过程理论建立了具有有限缓存的开排队网模型描述系统行为;提出了改进的状态空间分解法,并建立了节点状态空间模型,分析多类非同等并行自动导引小车与前后制造单元的耦合效应,并根据Markov过程理论分析节点状态转移规律;构造了迭代算法,并给出系统各项性能指标的计算方法。最后,设计了实验案例,与已有文献及仿真实验的结果进行对比,验证了改进状态空间分解法的精确性,并对系统性能指标进行了敏感性分析。

关键词: 自动导引小车, 柔性流水车间, 排队网, 状态空间分解法