控制工程 ›› 2019, Vol. 26 ›› Issue (12): 2225-2230.

• • 上一篇    下一篇

改进的EDA求解绿色可重入作业车间调度问题

  

  • 出版日期:2019-12-20 发布日期:2023-11-29

Improved EDA Solving Green Reentrant Job Shop Scheduling Problem

  • Online:2019-12-20 Published:2023-11-29

摘要: 针对生产过程对环境的巨大影响,提出一种基于贝叶斯统计的混合分布估计算法(Hybrid Bayesian-Statistical-Based Estimation of Distribution Algorithm, HBEDA)求解带交货期的多目标可重入作业车间绿色调度问题(Multi-objective Re-entrant Job Shop Green Scheduling Problem with Due Date, MRJSGSP_DD),实现对最大延迟(Maximum Tardiness, MT)和总能量消耗(Total Energy Consumption, TEC)的最小化。首先,在算法初始化阶段,产生一组随机种群,保证种群的随机性和多样性,并构造出非支配解集。其次,引入基于贝叶斯统计的混合分布估计算法构造概率模型,该模型能够学习到工件之间序的关系,增强了算法的全局搜索能力。最后,利用该问题的特性,设计了一种增强型的局部搜索的方法。仿真实验和算法对比验证了该算法的有效性。

关键词:

贝叶斯网络, 多目标, 绿色调度, 局部搜索, 作业车间调度

Abstract: In view of the huge impact to the environment in the production process, a HBEDA for solving the MRJSGSP_DD is put forward to achieve minimization of MT and TEC. Firstly, in the initialization phase of the algorithm, a group of random population is generated to guarantee the randomness and diversity of the population, and the non-dominated set is constructed. Then, HBEDA is introduced to construct the probabilistic model. The model can learn the relation between the orders of jobs and enhance the global searching ability of the algorithm. Finally, by using the characteristics of the problem, an enhanced local search method is designed. The effectiveness of the algorithm is verified by simulation and comparison.

Key words:

Bayesian network, multi-objective, green scheduling, local search, job-shop scheduling