(197b) Optimal production scheduling with interactive decision-making

In this paper, we present an optimization platform developed for production scheduling of continuous processes with special emphasis on utility systems. In large manufacturing units, commercial facilities and campuses, centralized utility plants meet a varying cooling, heating, and even electricity demands. The need to reduce environmental footprint of such plants along with the imperative for high efficiency mandate the use of smart and fast scheduling to achieve optimal operations. We underscore the key challenges in the deployment of such solutions specifically the need for accurate modeling of units, incorporating unit maintenance, and building operator confidence. The problem is formulated as mixed-integer non-linear program incorporating resource storage and parametric-first principle models. Our in-house optimization algorithm, based on Bender’s decomposition, is deployed as an On-demand cloud service for fast and scalable solution generation. We demonstrate the use of interactive optimization interfaces allowing operators to easily reschedule units, estimate the cost of their actions, and provide better forecasts for demands