In chemical production scheduling, we seek to allocate limited resources to competing tasks over time. The key decisions include the selection and sizing of tasks (batches), the assignment of tasks to equipment units, and the timing of tasks. Production planning and scheduling problems arise in many areas, from basic chemicals and consumer products to pharmaceuticals and specialty chemicals.
We classify chemical production scheduling problems by introducing triplet α/β/γ, where
α denotes the production environment. Chemical production environments are classified into three categories, namely network, sequential, and hybrid. Sequential environment can be further classified into multistage and multipurpose plants.
β denotes the processing restrictions and characteristics, such as setups, changeovers, storage constraints, material transfers. Note that multiple restrictions and characteristics can be present simultaneously in a problem.
γ denotes the objective function, including both minimization (makespan, cost, tardiness, etc.) and maximization (profit, production, etc.) objectives.
The goal of our research is to address problems of industrial size by formulating mixed-integer programming (MIP) models that can address the limitations of existing models and developing advanced solution strategies. Furthermore, we aim to understand the characteristics of online scheduling, where the schedule is constantly updated by repeatedly solving the problem, whenever a trigger event occurs or as the scheduling horizon advances. Specifically, we are investigating how different attributes of each open loop schedule affect the overall quality of the closed loop solution.