Abstract: Aiming at the prediction problem of production logistics bottleneck shifting under uncertain environment, a closed-loop prediction method of bottleneck was proposed based on bottleneck polymorphism. The concepts of sensitive bottleneck, obtuse bottleneck, growing bottleneck and decreasing bottleneck were formulated to describe the logistics bottleneck polymorphism of manufacturing system qualitatively. Based on bottleneck index, mathematical model of bottleneck shifting index was established by taking time and quality of manufacturing unit as parameters. Decision mechanism of bottleneck polymorphism was constructed to realize bottleneck polymorphism description on the basis of bottleneck shifting index quantitatively, which provided technological support for studying on bottleneck shifting rule. The bottleneck closed-loop prediction method which was composed of data statistic analysis, bottleneck prediction mechanism, bottleneck polymorphism analysis and bottleneck index iterative prediction was founded to realize dynamic and continuous prediction of production logistics bottleneck. An example of dynamically forecasting and monitoring the production logistics bottleneck in an automobile assembly shop was given to verify the effectiveness and accuracy of proposed prediction method.

Key words: production logistics, bottleneck polymorphism, bottleneck shifting index, bottleneck closed-loop prediction

摘要： 针对不确定性环境下的生产物流瓶颈漂移预测问题，提出一种基于瓶颈多态性的瓶颈闭环预测方法。建立灵敏型瓶颈、迟钝型瓶颈、渐向型瓶颈和渐离型瓶颈的概念，以定性描述制造系统的物流瓶颈多态性；以瓶颈指数为基础，以制造单元时间和质量为参数，构建瓶颈漂移指数数学模型，并以此为依据构建瓶颈多态性判定机制，以实现瓶颈多态性的定量描述，进而为瓶颈漂移规律的研究提供技术支持；构建包含数据统计分析、瓶颈预测机制、瓶颈多态性分析和瓶颈指数迭代预测四项内容的瓶颈闭环预测方法，以实现生产物流瓶颈的动态连续预测。通过对某汽车装配车间生产物流瓶颈的动态预测和监控，验证了该方法的有效性和准确性。

关键词: 生产物流, 瓶颈多态性, 瓶颈漂移指数, 瓶颈闭环预测