Application of Interval Predictor Model Into Model Predictive Control

In this paper, interval prediction model is studied for model predictive control (MPC) strategy with unknown but bounded noise. After introducing the family of models and some basic information, some computational results are presented to construct interval predictor model, using linear regression structure whose regression parameters are included in a sphere parameter set. A size measure is used to scale the average amplitude of the predictor interval, then one optimal model that minimizes this size measure is efficiently computed by solving a linear programming problem. The active set approach is applied to solve the linear programming problem, and based on these optimization variables, the predictor interval of the considered model with sphere parameter set can be directly constructed. As for choosing a fixed non-negative number in our given size measure, a better choice is proposed by using the Karush-Kuhn-Tucker (KKT) optimality conditions. In order to apply interval prediction model into model predictive control, the midpoint of that interval is substituted in a quadratic optimization problem with inequality constrained condition to obtain the optimal control input. After formulating it as a standard quadratic optimization and deriving its dual form, the Gauss-Seidel algorithm is applied to solve the dual problem and convergence of Gauss-Seidel algorithm is provided too. Finally simulation examples confirm our theoretical results.

Extending the Scope of Robust Quadratic Optimization

We derive computationally tractable formulations of the robust counterparts of convex quadratic and conic quadratic constraints that are concave in matrix-valued uncertain parameters. We do this for a broad range of uncertainty sets. Our results provide extensions to known results from the literature. We also consider hard quadratic constraints: those that are convex in uncertain matrix-valued parameters. For the robust counterpart of such constraints, we derive inner and outer tractable approximations. As an application, we show how to construct a natural uncertainty set based on a statistical confidence set around a sample mean vector and covariance matrix and use this to provide a tractable reformulation of the robust counterpart of an uncertain portfolio optimization problem. We also apply the results of this paper to norm approximation problems. Summary of Contribution: This paper develops new theoretical results and algorithms that extend the scope of a robust quadratic optimization problem. More specifically, we derive computationally tractable formulations of the robust counterparts of convex quadratic and conic quadratic constraints that are concave in matrix-valued uncertain parameters. We also consider hard quadratic constraints: those that are convex in uncertain matrix-valued parameters. For the robust counterpart of such constraints, we derive inner and outer tractable approximations.