Comprehensive study of decomposition effects on distributed output tracking of an integrated process over a wide operating range

A comprehensive study of plant decomposition effects is presented for distributed model predictive control (DMPC) of an integrated process system. Different decompositions are obtained via community detection and other methods. The closed-loop performance and computational efficiency of employing various decompositions for DMPC design are evaluated through tracking outputs in different tracking zones corresponding to desired operating conditions. Different levels of communication and cooperation between local controllers, levels of system uncertainty, and dynamic optimization platforms are considered. The results are analyzed to determine the most suitable method for decomposition of the studied integrated process.