Continuous Production\nScheduling MILP Formulations\nUsing Record Keeping Variables

Most solution methods for mixed-integer linear programming\n(MILP)\nproduction scheduling models have been developed for batch processes.\nIn this paper, we employ integer variables, referred to as record\nkeeping variables (RKVs), into discrete-time continuous production\nscheduling MILP models that facilitate efficient branching and lead\nto substantial reductions in solution time. We first introduce different\ntypes of RKVs and determine which class of RKVs is the most effective.\nSecond, we explore branching priorities and demonstrate that prioritizing\nbranching on RKVs, relative to other binary variables, leads to further\ncomputational improvements. Next, we analyze system attributes, such\nas task and unit utilization, to determine if prioritizing branching\non specific RKVs leads to additional computational enhancements. Our\ncomputational results show that the proposed reformulations, in combination\nwith implementing branching priorities, lead to significant computational\nimprovements of continuous production scheduling MILP models.