A State Space Model Predictive Control Method For Formation Adjustments Of Process Goose Queue

It is recognized that the Process Goose Queue (PGQ) approaches can effectively deal with decomposition and coordination optimization problems of complex industrial processes, enjoying easily modeling and fast convergence rates. However, a major defect associated with the exciting PGQ approaches is identified as the relatively slow formation adjustment speeds in the presence of disturbances caused by inherent limitations of the hierarchical tracking objective transmission.Motivated by this challenge, an ontology-based framework of state space model predictive control is created in this thesis. Initially, taking advantage of the correlating relationship associated with the PGQ systems, the state space models of individual PGQs are established along with the corresponding sub-space identification algorithms. Consequently, a state-space model predictive control strategy towards the formation adjustments of the PGQ systems is explicitly proposed. It reveals that this approach can help predict outputs of the higher layer PGQ based on the states of the PGQ interrupted so as to adjust the manipulated following goose positions of the both PGQ layers simultaneously, eliminating impacts of the interrupted PGQ on optimization targets of the multi-level PGQ systems quickly.The TE process is employed as the case studies, demonstrating that the proposed approach enjoys strong abilities in disturbance rejection, as well as smooth adjustments and fast dynamic responses.