基于制造执行系统数据采集的工时预测与进化

Man-hour forecasting and evolution based on MES data collection

为提高定额工时的准确性,提出一种基于制造执行系统数据采集的工时预测与进化技术。给出了基于制造执行系统加工历史数据的实做工时统计方法,对工时影响因素进行分析并建立了工时预测神经网络。建立了工时进化过程模型并对进化过程进行描述,利用实做工时持续对定额工时与工时预测神经网络进行优化与训练完成工时的进化,通过进化使定额工时及其预测神经网络符合实际生产。在某航天企业对该方法的效果与可行性进行了验证。

To improve the accuracy of rated man-hours, a man-hour evolution method based on Manufacturing Execution System （MES） data collection was put forward. Actual work time statistic method based on MES historian data was given, man-hour influencing factors were analyzed and neural network was set up. Man-hour evolution process model was established and the evolutionary process was described. The evolution process was carried out by using real work time to revise man-hour and its neural network constantly, and they were kept consistent with the actual manufacturing situation by means of evolution. The feasibility of proposed method was validated in an aerospace enterprise.