扰动累积下基于数字孪生的车间重调度

针对隐性扰动累积下车间重调度问题，搭建车间重调度决策服务，实现数字孪生驱动的扰动累积下的车间重调度。提出一种调度参数更新的方式，采用随机概率分布来描述调度参数的分布，提升调度参数的准确性；利用孪生网络搭建隐性扰动检测模型，以实时数据为输入，实现重调度的启动时刻的判定；从历史调度数据中提取用于调度知识挖掘的样本数据，利用伪孪生网络获取工序和机器的状态数据的映射关系作为调度规则，用于车间的重调度。仿真实验证明了所提数字孪生驱动的重调度模式的可行性。

Job Shop Rescheduling Under Recessive Disturbance Based on Digital Twin

A new shop rescheduling model driven by digital twin is proposed to solve the problems of disturbance cumulative rescheduling. A scheduling parameter updating method is proposed and a random probability distribution is used to describe the distribution of scheduling parameters to improve the accuracy of scheduling parameters. An implicit disturbance detection model is built based on Siamese Network using real-time data as input to realize the start time of rescheduling. The sample data for scheduling knowledge mining are extracted from the historical scheduling scenarios. Through the Pseudo-Siamese CNN, the mapping relationship between the Process state and machine state is obtained, which is applied to production online rescheduling. Simulation experiments show the feasibility of the proposed digital twin driven shop rescheduling model.