Achieving higher frequencies in large-scale nonlinear model predictive control

We present new insights into how to achieve higher frequencies in large-scale nonlinear predictive control using truncated-like schemes. The basic idea is that, instead of solving the full nonlinear optimization (NLO) problem at each sampling time, we solve a single, truncated quadratic optimization (QO) problem. We present conditions guaranteeing stability of the approximation error for truncated schemes using generalized equation concepts. In addition, we propose a preliminary scheme using an augmented Lagrangian reformulation of the NLO and projected successive overrelaxation to solve the underlying QO. This strategy enables early termination of the QO solution because it can perform linear algebra and active-set identification tasks simultaneously. A simple numerical case study is provided.