Abstract
Process scheduling is one key layer of decision hierarchy for process industries to optimize their production schedule in order to gain the long-term economic viability. Besides coordinating limited available resources and satisfying demands on production quantity, quality, and environmental restrictions, the challenge of process scheduling also lies in the treat of uncertainties when approaching the multistage adjustable robust optimization (ARO) of the scheduling problem. In this work, we solve the robust optimization problem for batch process scheduling under uncertainty by using piecewise linear decision rule (PLDR). Based on PLDR and the Dirichlet process Gaussian mixture model (DPGMM) which is for data-driven uncertainty set construction, we demonstrate with an industrial process case study that the combination of the data-driven uncertainty set and the piecewise linear decision rule is capable of generating usually better, at least as good, batch process scheduling optimization solutions in comparison to some conventional ARO approaches.
