Abstract: Aiming at the bottleneck identification problem in job-shop operation management of work station,a new Web-based manufacturing bottleneck identification method was presented.A job-shop network model was established according to multiple levels of production data such as equipment and tooling,process route,logistics path and product configuration.By expending bottleneck connotation,a bottleneck recognition method was proposed based on efficiency of network bottlenecks matrix.The comprehensive evaluation of the proposed method was taken into account combined with the network structure,network communication mechanisms and node self characteristics.Furthermore,the bottleneck index was characterized by using manufacturing load and network efficiency matrix,which overcame the deficiency that the bottleneck node identifier depended only on the adjacent node.An example for dynamic monitoring and forecasting the bottleneck in a mechanical and electrical products plant job shop was given to prove the validation and practicability of bottleneck．The result showed that the network model of manufacturing system was an effective way to achieve rapid manufacturing system assessment.

Key words: complex network, job-shop, bottleneck contribution, bottleneck identification

摘要： 针对作业车间管理层面的瓶颈识别,从制造系统复杂性与复杂网络相结合这一全新视角,提出基于网络特性的制造瓶颈识别方法。根据设备工装、工艺路线、物流路径和产品配置等多层次生产数据,建立了作业车间网络模型。对瓶颈内涵进行了扩充,提出基于网络瓶颈效率矩阵的瓶颈识别算法。综合考虑了网络结构、网络传播机制及节点自身特性的影响,利用节点的制造负载和节点间的网络瓶颈效率矩阵表征节点的瓶颈程度,克服了其他算法中瓶颈节点识别只依赖于邻接节点的不足。通过对某机电产品企业车间生产瓶颈的动态监控和预测,验证了该瓶颈识别方法的有效性和准确性,也表明制造系统的网络模型是实现制造系统快速评估的有效方法。

关键词: 复杂网络, 作业车间, 瓶颈贡献度, 瓶颈识别