A Distributed State-Space Model Predictive Control Based Approach To Multi-Level Process Goose Queue Adjustments

PGQ (Process Goose Queue) is a novel approach to deal with the decomposition and coordination optimization of complex industrial processes where an industrial process is decomposed according to natural goose queue flight mechanism. Associated PGQs exchange information to transfer the optimal value in multi-lever PGQ. However, when the multi-player PGQ is affected by disturbances, a sick PGQ appears and it takes much time to get to the steady state again, which hinders the implementation of the overall optimization objectives.Motivated by this challenge, a hierarchically distributed model predictive control algorithm is proposed in this paper, which decomposes the overall optimal objective into each PGQ. It's discovered that the state of the PGQ is only affected by the lower sub Goose. Based on the cooperated distributed model predicted control and taking advantage of the information flow from lower PGQ to higher PGQ, the optimization problem of the multi-level PGQ is solved hierarchically, which avoids the iterative process and greatly reduces the burden of communication of the algorithm, achieving the performance close to the central predictive control.The method was applied to TE process, the experiment results show that the proposed approach demonstrates strong abilities in fast dynamic responses, disturbance rejection and smooth adjustments.