Deterministic Global Nonlinear Model Predictive Control with Neural Networks Embedded

Doncevic, Danimir; Caspari, Adrian; Vaupel, Yannic; Schweidtmann, Artur M.; Mitsos, Alexander; Schäfer, Pascal

doi:10.1016/j.ifacol.2020.12.1207

Journal Article/Contribution to a conference proceedings

FZJ-2021-03179

Deterministic Global Nonlinear Model Predictive Control with Neural Networks Embedded

Doncevic, D.FZJ*RWTH* ; Schweidtmann, A. M.RWTH* ; Vaupel, Y.RWTH* ; Schäfer, P.RWTH* ; Caspari, A.RWTH* ; Mitsos, A. (Corresponding author)FZJ*RWTH*

2020
IFAC Laxenburg

1st Virtual IFAC World Congress, online, Germany, 11 Jul 2020 - 17 Jul 2020 IFAC-PapersOnLine 53(2), 5273 - 5278 (2020) [10.1016/j.ifacol.2020.12.1207]

This record in other databases:

Please use a persistent id in citations: http://hdl.handle.net/2128/28412 doi:10.1016/j.ifacol.2020.12.1207

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.

Classification: