Feedforward control of a channel flow based on a discretized port-Hamiltonian model

Systems of conservation laws can be modeled (including dissipation) in an elegant, physically insightful way within the port-Hamiltonian framework. A structure-preserving discretization renders the partial differential equations ordinary ones. In this paper, we show how the structure of the lumped-parameter state representation for two conservation laws On a One-dimensional. spatial domain can he exploited to easily formulate different (inverse) models. Based thereon, a simple modular procedure for feedforward controller design is developed, using known results from the dynamic inversion of nonminimilm-phase systems. The example of the shallow water equations serves to illustrate the design steps and to present simulation results. (C) 2015, ILAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.