一种改进的顺序型多目的间歇工厂生产调度的MILP模型

Modified MILP model for scheduling of sequential multipurpose batch plants

分析了顺序型多目的工厂间歇生产的特点,建立了一种新的混合整数线性规划(MILP)模型.该模型以操作活动序列作为建模的依据,将整个间歇生产过程分解成若干个子系统.在对各子系统建模的基础上,实现了顺序型多目的工厂间歇调度的全局性优化.并通过算例验证了该模型求解结果的正确性和可行性.然后将新模型运用到更大规模的调度问题上,均求解出了有效的调度方案.

Sequential multipurpose batch processes are common in many fine chemical manufacturing plants, such as traditional paint, food, pharmaceuticals, and specialty chemicals plants. The investigation in this paper includes analyzing the operational features of sequential multipurpose batch plants, decomposing the overall batch process spatially into several sub-systems, and presenting a modified MILP formulation for the scheduling of sequential multipurpose batch plants with continuous-time approaches. As the bi-index binary variables are substituted for the tri-index binary variables in the MILP model, the number of discrete variables decreases substantially. The new model is effective in solving large scale scheduling problems. In case studies, the proposed model was compared with the previous model presented by Voudouris and Grossmann. Moreover, several large scale scheduling problems were solved successfully with the proposed modeling approaches. These case studies prove the validity of the model and the practicability of the proposed approaches.