Abstract: For an aero-engine assembly shop with missing parts,the self-evolution problem of knowledgeable manufacturing systems was researched.Rolling horizon method was combined with hybrid event-driven and periodic-driven self-evolution mechanism to implement the workshop's self-evolution.Based on the characteristics of  production,a rolling rule was presented to select operations into the rolling window at each decision moment,a mathematical model of the static decision problem at each decision point was built,and an algorithm was developed to solve the self-evolution problem.For the model,a Genetic-Variable Neighborhood Search (GVNS) algorithm was designed.The simulations were conducted to analyze the performance of the algorithm.The experimental data showed that the self-evolution had an important role in improving the system production performance.

Key words: knowledgeable manufacturing system, self-evolution, aero-engine assembly shop, rolling horizon, genetic-variable neighborhood search 

摘要： 针对存在缺件情况的航空发动机装配车间,研究了知识化制造系统自进化问题。采用事件和周期混合驱动型自进化机制,结合滚动时域法实现车间自进化。基于生产特点提出一种滚动规则,用于各决策时刻选取工序进入滚动窗口,建立了每个决策时刻系统中静态决策问题的数学模型,并给出了自进化问题的求解算法。针对模型设计了一种遗传—变邻域搜索算法进行求解。通过仿真实例对算法的性能进行了分析。实验数据表明,自进化在提升系统生产性能方面发挥了重要的作用。

关键词: 知识化制造系统, 自进化, 航空发动机装配车间, 滚动时域, 遗传&mdash, 变邻域搜索算法