基于预测机制的装配作业重调度问题建模优化

飞机移动装配中因作业质量问题增加的修复作业会打乱装配计划并造成经济损失,为解决该问题提出了一种预测-重调度形式的闭环框架,该框架的前端利用质量相关部件衰退以及作业质量特性偏差的历史数据训练了作业质量预测模型,基于预测结果,重点针对后端建立了装配作业重调度模型并设计了改进型免疫算法(I-I-A)用于生成新的装配计划。数值实验部分从不同角度验证了I-I-A的有效性,同时也对所提闭环框架的性能优劣进行了对比分析。

Modeling and Optimization of Job Assembly Rescheduling Problem Based on Job Quality Prediction Mechanism

Repair jobs resulting from poor assembly quality of jobs in the aircraft assembly process can disturb assembly scheduling plans and cause economic losses. This paper proposes a prediction-rescheduling closed-loop framework in order to solve this scheduling problem. In the front part of this framework, job quality prediction models are trained using historical data of quality-related components in deteriorations and quality characteristic deviations of jobs. Based on the prediction results of jobs, a rescheduling model is established in the later part of this framework and an improved-immune-algorithm (I-I-A) is designed to generate a new scheduling plan for the assembly line. The effectiveness of the I-I-A is verified from different aspects and the advantages and disadvantages of the performance of the closed-loop framework suggested in this paper is analyzed in comparison with other frameworks.