Abstract: Asynchronous production lines without buffers are composed of multiple machines with different cycle times,widely applied in various manufacturing fields,such as automotive and aero-engine manufacturing.The two features of asynchronization and no buffer  impact the characterization of the state space,and consequently affecting the performance modeling of the production lines.The impact of machine random failures and cycle differences on the system performance was analyzed to study the modeling and performance analysis of asynchronous production lines.The activities of machines and jobs in the production line were modeled by an activity network.An analytical model for performance analysis was established to describe the functional relationship between completion time and random variables,including holding time,starvation time and blockage time.Furthermore,the correlation between performance measures and random variables was quantitatively analyzed,and an iterative algorithm was adopted to evaluate performance measures,such as production rate,probabilities of machine starvation and blockage.Finally,the analytical model was validated by simulation experiments.

Key words: production lines without buffers, cycle differences, production line modeling, analytical method, performance analysis

摘要： 无缓冲异步流水线由多台加工周期不同的机器串联组成,广泛应用于汽车、航空发动机等制造领域。无缓冲与异步两大特征直接影响系统状态空间刻画,进而影响产线性能建模。面向无缓冲异步流水线,分析机器随机故障、节拍差异等因素对产线性能的影响,开展产线建模与性能分析研究。从机器维和工件维等两个维度出发,利用活动网络图描述了流水线中工件在机器上的活动状态,刻画了工件保持时间、机器饥饿时间、机器阻塞时间等随机变量与工件完工时间的函数关系,建立了产线性能分析的解析模型。在此基础上,定量分析了系统性能指标与工件保持时间等随机变量之间的关联关系,采用迭代算法求解了生产率、饥饿率和阻塞率等系统性能指标,并通过仿真实验验证了解析模型的有效性。

关键词: 无缓冲流水线, 节拍差异, 产线建模, 解析方法, 性能分析