Decision support system for a heat-recovery section with equipment degradation

Abstract

In the framework of Industry 4.0, decision support systems (DSS) are an essential part in the process of converting information into managerial actions. This paper proposes a specific DSS to improve the daily operation of an industrial heat-recovery section (a network of heat exchangers) in a fiber-production factory. In this process, the aim is to optimize the resource utilization in real time while satisfying a set of production constraints. The operational decisions to be taken by the network operators is to set up the heat-source allocation to heat exchangers. Furthermore, the heat transfer decreases over time due to fouling in the exchangers, so an additional decision to take is which exchanger to clean and when. The proposed model-based DSS builds upon a rigorous mathematical representation of the network, integrating continuous operation with the discrete decisions on maintenance. Then, a mixed-integer nonlinear optimization, solved in real time, drives the analysis and choice phases to fulfill product specifications according to an economic criterion. In this way, the proposed DSS not only provides the user with a right allocation of heat sources to exchangers, but also suggests which of them are potentially beneficial to be cleaned.