首页|基于改进进化算法的中药提取车间批调度问题研究

基于改进进化算法的中药提取车间批调度问题研究

扫码查看
现阶段中药提取车间生产周期长、物料浪费严重,实际生产时其生产相关时间和物料损失之间存在着复杂的制约关系,导致调度难度高,排产效率低.针对此紧密耦合关系,以最小化完工时间和最小化物料总损失量为优化目标,建立了混合整数规划模型,提出了一种改进的进化算法.算法在初始化种群阶段提出了一种基于概率矩阵的初始化策略,以减少算法搜索范围;在进化阶段引入精英保留策略,以优化种群质量;在变异阶段设计了 一种邻域搜索策略,保证算法全域搜索能力.通过田口实验设计方法确定算法参数组合,利用策略有效性实验验证算法有效性.与NSGA Ⅱ和SPEA2进行实验对比,验证了算法的优越性、稳定性与更优的求解效率.最后与实际生产方案比对,结果表明利用改进算法所得调度结果在完工时间与物料损失量方面均明显优于现有方案.
Research on Batch Scheduling Problem of Traditional Chinese Medicine Extraction Workshop Based on Improved Evolutionary Algorithm
At present,the traditional Chinese medicine extraction workshop has a long production cycle and severe material waste.In actual production,there is a complex constraint relationship between production related time and material loss,resulting in high scheduling difficulty and low production efficiency.In view of this close coupling relationship,a mixed integer programming model was established to minimize the completion time and the total material loss.An improved evolutionary algorithm was proposed.The algorithm proposed an initialization strategy based on probability matrix during the initialization population stage to reduce the search range of the algorithm.Elite retention strategies ware introduced in the evolutionary stage to optimize population quality.A neighborhood search strategy was designed during the mutation stage to ensure the algorithm's global search ability.The combination of algorithm parameters was determined through the Taguchi experimental design method.The effectiveness of the algorithm was verified through strategy effectiveness experiments.Compared with NSGA Ⅱ and SPEA2 through experiments,the superiority,stability,and better solving efficiency of the algorithm were verified.Finally,compared with the actual production plan,the results show that the scheduling results obtained using the improved algorithm are significantly better than the existing plan in terms of completion time and material loss.

batch schedulingmulti-objective optimizationevolutionary algorithmsdrug extraction

罗亚波、王洲旭

展开 >

武汉理工大学机电工程学院,湖北武汉 430070

批调度 多目标优化 进化算法 药物提取

国家自然科学基金

51875430

2024

工业工程与管理
上海交通大学

工业工程与管理

CSTPCD北大核心
影响因子:0.763
ISSN:1007-5429
年,卷(期):2024.29(4)
  • 3