Deterministic global nonlinear model predictive control with neural networks embedded

Journal article (2020)

Authors

Danimir T. Doncevic Rheinisch-Westfälische Technische Hochschule, Forschungszentrum Jülich
A.M. Schweidtmann Rheinisch-Westfälische Technische Hochschule
Yannic Vaupel Rheinisch-Westfälische Technische Hochschule
Pascal Schäfer Rheinisch-Westfälische Technische Hochschule
Adrian Caspari Rheinisch-Westfälische Technische Hochschule
Alexander Mitsos Rheinisch-Westfälische Technische Hochschule, Forschungszentrum Jülich, JARA Center for Simulation and Data Science (CSD)
Affiliation
External organisation
Global optimization Recurrent neural networks Model predictive and optimization-based control Neural networks in process control Nonlinear process control
More Info
expand_more

Published Date

2020

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Affiliation

External organisation

Abstract

Nonlinear model predictive control requires the solution of nonlinear programs with potentially multiple local solutions. Here, deterministic global optimization can guarantee to find a global optimum. However, its application is currently severely limited by computational cost and requires further developments in problem formulation, optimization solvers, and computing architectures. In this work, we propose a reduced-space formulation for the global optimization of problems with recurrent neural networks (RNN) embedded, based on our recent work on feed-forward artificial neural networks embedded. The method reduces the dimensionality of the optimization problem significantly, lowering the computational cost. We implement the NMPC problem in our open-source solver MAiNGO and solve it using parallel computing on 40 cores. We demonstrate real-time capability for the illustrative van de Vusse CSTR case study. We further propose two alternatives to reduce computational time: i) reformulate the RNN model by exposing a selected state variable to the optimizer; ii) replace the RNN with a neural multi-model. In our numerical case studies each proposal results in a reduction of computational time by an order of magnitude.