Large-scale Optimization Formulations and Strategies for Nonlinear Model Predictive Control

Abstract

Concepts, algorithms and modeling platforms are described for the realization of nonlinear model predictive control (NMPC) using nonlinear programming (NLP). These allow the incorporation of predictive nonlinear dynamic models that lead to high performance control, estimation and optimal operation. As essential background, we first review NLP formulations that guarantee properties for nominal and robust stability. In addition, fast algorithms for NMPC are described to deal with sensitivity-based solutions. To implement these results in practice, we discuss a recently developed Python-based modeling platform that is tailored to deal with dynamic optimization strategies for state estimation and nonlinear control. Finally, we present ongoing work on parallel decomposition strategies for large dynamic NLPs, which will lead to further gains in computational speed as well as robust performance gains.