Abstract: Stochastic production system containing multiple manufacturing cells with Automatic Guided Vehicle (AGV) has become a typical manufacturing system structure for customized enterprises.To analyze the performance of manufacturing system with AGV and nested tracks,the uncertainty of manufacturing system with the stochastic path of AGV were considered,and the operation of system was described by queuing network model with finite buffer.The traditional state space decomposition method was expanded to establish state model of each node according to the relation between cells and stochastic-path AGV,and pure discontinuous Markov process was used to analyze the state transferring discipline between each node.Iteration algorithm was constructed to calculate the stable probability of node state.A simulation model was established,and the simulation experiment was designed and conducted to testify the validity of the proposed state space decomposition method.The sensitivity of system performance was also analyzed.

Key words: automatic guided vehicle, stochastic path, multiple nested tracks, queuing network, state space decomposition

摘要： 为有效分析具有随机路径自动导引小车(AGV)且多车道嵌套制造系统的性能,考虑AGV状态相关随机运输路径和制造系统的不确定性,采用随机过程理论建立具有有限缓存的开排队网模型描述系统特征;扩展传统的状态空间分解法,依据随机路径AGV与前后制造单元的关联耦合效应建立各节点状态模型,通过纯不连续Markov过程理论分析节点状态转移规律;构建迭代算法求解节点状态的稳态概率,并给出系统各性能指标的计算方法;搭建了系统仿真模型,设计了实验案例,并对比排队网模型与仿真模型的实验结果,验证了状态空间分解法近似求解的有效性,同时对系统的性能指标进行了敏感度分析。

关键词: 自动化导航小车, 随机路径, 多车道嵌套, 排队网, 状态空间分解法