Nonlinear Model Predictive Control Based on Operating-Trajectory Model

ZHU Yu-cai
DOI: https://doi.org/10.3969/j.issn.1007-0249.2009.01.012
2009-01-01
Abstract:Nonlinear model predictive control based on operating-trajectory model is proposed.In nonlinear process identification,the so-called operating-trajectory model approach is used.Firstly,typical working-points are selected based on operating-trajectory of scheduling variable;Secondly,linear models are identified using data sets at various working-points exclusive transition data;Finally,the LPV model is identified by interpolating the linear models using total data.Further,nonlinear model predictive control based on operating-trajectory model is proposed.The control action is computed via multi-step linearization method of nonlinear optimization problem.The method uses low cost tests and keeps better control performance.Simulation verifies the effectiveness of the method.
What problem does this paper attempt to address?