中文摘要：

针对扰动环境下作业车间多瓶颈识别困难、瓶颈漂移后的瓶颈识别缺乏全局性和实效性这一问题,构建了基于网络特性的多瓶颈动态识别方法。首先,根据设备工装、工艺路线、物流路径以及产品配置等多层次生产数据,构造作业车间网络模型;其次,建立作业车间网络动力学方程,获取扰动因素流转的判定依据。对瓶颈内涵进行扩充,综合考虑节点自身动力学特性、节点间拓扑耦合影响机理及扰动在生产网络中的传播机制,建立基于耦合映射格子（CML）的瓶颈识别算法,实现作业车间瓶颈的量化描述和连续预测;最后,对某机电企业作业车间进行瓶颈的动态监控和预测。结果表明：在扰动环境下,CML模型能够较好地预测各工作站瓶颈度走势,其中工作站R1平均瓶颈度为1.12,瓶颈持续时间长达40h;工作站R3的平均瓶颈度为1.05,瓶颈持续时间为10h;工作站R1、R3首先成为系统的瓶颈,随着加工进度的推移,工作站R1和R24交替成为系统瓶颈。研究结果与该企业实际情况具有很好的一致性,验证了该方法的有效性和准确性。

英文摘要：

Aiming at the issues that the bottleneck identification is difficult and the bottleneck identification is not global and effective after bottleneck drift for job-shop network in disturbance environment, a new web-based manufacturing multi-bottleneck identification method is presented. A network model of job-shop is established according to multiple levels of production data, such as equipment and tooling, process route, logistics path, product configuration, etc. The kinetic equations of job-shop network are established, and the criterion of transfer of disturbance factors is thus obtained. Expending bottleneck connotation, a bottleneck identification strategy is proposed based on CML method. By comprehensively considering the characteristics of dynamic node itself, network topological structure and propagation mechanism of disturbance in the network, continuous quantitative description and prediction of job-shop bottlenecks are realized. An example for dynamically monitoring and forecasting the bottleneck in a job shop verifies the validation and practicability of the proposed method. The results show that in disturbance environment, CML model can better predict the trend of the bottleneck degree of the workstations. The average bottleneck degree of workstation R1 is 1.12 and the bottleneck lasts for 40 h, and the average bottleneck degree of workstation R3 is 1.05 and the bottleneck lasts for 10 h. Workstations R1 and R3 become the first bottleneck of the system, along with the progress of the process workstations R1 and R24 alternately become the system bottleneck.