Mixed Integer Optimal Control of Systems with Typical Nonlinear Components

张聚,李平
DOI: https://doi.org/10.3785/j.issn.1008-973x.2004.11.005
2004-01-01
Abstract:By equivalently transforming propositional logic into linear inequalities involving auxiliary logical and continuous variables, a general linear system model (GLSM) was established to describe accurately a wide range of nonlinear systems with typical non-smooth and even discontinuous nonlinear components. A GLSM based mixed integer predictive control (MIPC) approach with dynamic predictive horizon for optimal regulating and tracking problem of a class of nonlinear systems with typical nonlinear components was studied. A standard mixed integer quadratic programming (MIQP) problem was solved at every time instant when MIPC was implemented. The already developed branch and bound (B B) based standard MIQP solver was used for the numerical simulations, which showed that the approach developed was suitable and effective for optimal control of systems with typical nonlinear components.
What problem does this paper attempt to address?