Authors: Yan Zhang
Addresses: Department of Electrical Engineering and Automation, Shanghai Maritime University, 1550 Pudong Road, Shanghai 200135, China
Abstract: A networked MPC scheme based on dynamic integrated system optimisation and parameter estimation (DISOPE) is proposed for online optimisation and control of plant-wide processes. Under network environment, online optimisation of the whole system is decomposed into that of several small-scale subsystems, and the connectivity of communication network is assumed to be sufficient for each subsystem to exchange information with other subsystems. An iterative algorithm for DISOPE-based networked MPC is developed in spite of the model-reality differences by iterating on modified linear optimal control problems on the basis of estimating parameters at each iteration. It can be proved that the converged solution of the iterative scheme satisfies the optimality conditions of the original non-linear optimisation problem under some assumptions via computational convergence analysis. The simulation results of the fossil fuel power unit (FFPU) are illustrated to verify the effectiveness and practicability of the presented algorithm.
Keywords: plant-wide processes; dynamic integrated system optimisation; parameter estimation; DISOPE; networked MPC; linear optimal control; simulation; fossil fuel power units; model predictive control; power plants.
International Journal of Modelling, Identification and Control, 2011 Vol.14 No.3, pp.178 - 188
Published online: 23 Sep 2011 *Full-text access for editors Access for subscribers Purchase this article Comment on this article