10th IFAC Symposium

on Advanced Control of Chemical Processes

July 25 - 27, 2018 Shenyang, China


Towards industrial robust nonlinear model predictive control


Sergio Lucia. Internet of Things for Smart Buildings, Technische Universität Berlin & Einstein Center Digital Future, Berlin, Germany.

Sankaranarayanan Subramanian, Sebastian Engell. Process Dynamics and Operations Group, Technische Universität Dortmund, Dortmund, Germany.


Sergio Lucia (TU Berlin & Einstein Center Digital Future)

Sankaranarayanan Subramanian (TU Dortmund)

Sebastian Engell (TU Dortmund)


Half a day.


Model Predictive Control (MPC) has become one of the most popular control techniques in the process industry mainly because of its ability to deal with multipleinput-multiple-output plants and with constraints including also economic objectives. However, its performance can deteriorate in the presence of model uncertainties and disturbances. In the last years, the development of robust MPC techniques has been widely discussed, but these were rarely applied in practice due to their conservativeness or their computational complexity.

This workshop presents multi-stage nonlinear model predictive control (multi-stage NMPC) as a promising non-conservative robust NMPC control scheme, which is applicable in real-time. The approach is based on the representation of the evolution of the uncertainty by a scenario tree. It leads to non-conservative robust control of the plant because it takes into account explicitly that new information (usually present as measurements) will become available at future time steps and that the future control inputs can be adapted accordingly, acting as recourse variables.

The workshop starts with an introduction to the state of the art in robust nonlinear model predictive control, including min-max, tube-based and multi-scenario approaches. After illustrating its limitations, we will present the main idea of multistage NMPC, its mathematical formulation and theoretical properties as well as simulation results for an industrial case study.

In the second part of the workshop, we focus on output feedback and dual/adaptive formulations of multi-stage NMPC. We discuss the output feedback formulations using the multi-stage NMPC framework that provides robustness against estimation errors in addition to plant-model mismatches. Adaptive multi-stage NMPC that adapts the scenario-tree for time-varying and time-invariant uncertainties will be presented along with dual-control formulations that take into account the balance between explorative and optimal actions.

We will also present the tool do-mpc as a way to easily formulate and efficiently solve multi-stage NMPC problems.

Agenda (4h 00min)

15 min


Why do we need robust nonlinear model predictive control?

(Sebastian Engell, TU Dortmund)

30 min


Dealing with uncertainty: existing robust NMPC approaches

(Sergio Lucia, TU Berlin & Einstein Center Digital Future)

45 min


Handling uncertainty with multi-stage NMPC

(Sergio Lucia, TU Berlin & Einstein Center Digital Future)

30 min


Output feedback multi-stage NMPC

(Sankaranarayanan Subramanian, TU Dortmund)

15 min

Coffee break

30 min


Adaptive/Dual control in the framework of multi-stage NMPC

(Sankaranarayanan Subramanian, TU Dortmund)

30 min


Using do-mpc: a modular tool for efficient robust NMPC

(Sergio Lucia, TU Berlin & Einstein Center Digital Future)

30 min


Application of multi-stage NMPC to a reactive distillation column

(Sebastian Engell / Daniel Haßkerl, TU Dortmund)

15 min


Future challenges and discussion