A Data-Driven Model Predictive Control Approach to Lean NOx Trap Regeneration

2018 ◽  
Vol 141 (1) ◽  
Author(s):  
Milad Karimshoushtari ◽  
Carlo Novara
Keyword(s):  
Model Predictive Control ◽  
Predictive Control ◽  
Prediction Models ◽  
Low Complexity ◽  
Data Driven ◽  
Lean Nox Trap ◽  
Nox Trap ◽  
Control Approach ◽  
Highly Nonlinear ◽  
Lean Nox

Lean NOx trap (LNT) is one of the most effective after-treatment technologies used to reduce NOx emissions of diesel engines. One relevant problem in this context is LNT regeneration timing control. This problem is indeed difficult due to the fact that LNTs are highly nonlinear systems, involving complex physical/chemical processes, that are hard to model. In this paper, a novel approach for regeneration timing of LNTs is proposed, allowing us to overcome these issues. This approach, named data-driven model predictive control (D2-MPC), does not require a physical model of the engine/trap system but is based on low-complexity polynomial prediction models, directly identified from data. The regeneration timing is computed through an optimization algorithm, which uses the identified models to predict the LNT behavior. Two D2-MPC strategies are proposed, and tested in a co-simulation study, where the plant is represented by a detailed LNT model, built using the well-known commercial tool AMEsim, and the controller is implemented in matlab/simulink.

Two-Level Nonlinear Model Predictive Control for Lean NOx Trap Regenerations

2010 ◽  
Vol 132 (4) ◽  
Author(s):  
Ming-Feng Hsieh ◽  
Junmin Wang ◽  
Marcello Canova
Keyword(s):  
Diesel Engine ◽  
Model Predictive Control ◽  
Predictive Control ◽  
Nonlinear Model ◽  
Nox Reduction ◽  
Nonlinear Model Predictive Control ◽  
Lean Nox Trap ◽  
Nox Trap ◽  
Profile Control ◽  
Lean Nox

This paper describes a two-level nonlinear model predictive control (NMPC) scheme for diesel engine lean NOx trap (LNT) regeneration control. Based on the physical insights into the LNT operational characteristics, a two-level NMPC architecture with the higher-level for the regeneration timing control and the lower-level for the regeneration air to fuel ratio profile control is proposed. A physically based and experimentally validated nonlinear LNT dynamic model is employed to construct the NMPC control algorithms. The control objective is to minimize the fuel penalty induced by LNT regenerations while keeping the tailpipe NOx emissions below the regulations. Based on the physical insights into the LNT system dynamics, different choices of cost function were examined in terms of the impacts on fuel penalty and tailpipe NOx slip amount. The designed control system was evaluated on an experimentally validated vehicle simulator, cX-Emissions, with a 1.9 l diesel engine model through the FTP75 driving cycle. Compared with a conventional LNT control strategy, 31.9% of regeneration fuel penalty reduction was observed during a single regeneration. For the entire cold-start FTP75 test cycle, a 28.1% of tailpipe NOx reduction and 40.9% of fuel penalty reduction were achieved.

Probabilistic model predictive control for extended prediction horizons

2021 ◽  
Vol 69 (9) ◽  
pp. 759-770
Author(s):  
Tim Brüdigam ◽  
Johannes Teutsch ◽  
Dirk Wollherr ◽  
Marion Leibold ◽  
Martin Buss
Keyword(s):  
Model Predictive Control ◽  
Predictive Control ◽  
Prediction Models ◽  
Computational Effort ◽  
Probabilistic Constraints ◽  
Detailed Model ◽  
Control Approach ◽  
Proposed Model ◽  
Sampling Times

Abstract Detailed prediction models with robust constraints and small sampling times in Model Predictive Control yield conservative behavior and large computational effort, especially for longer prediction horizons. Here, we extend and combine previous Model Predictive Control methods that account for prediction uncertainty and reduce computational complexity. The proposed method uses robust constraints on a detailed model for short-term predictions, while probabilistic constraints are employed on a simplified model with increased sampling time for long-term predictions. The underlying methods are introduced before presenting the proposed Model Predictive Control approach. The advantages of the proposed method are shown in a mobile robot simulation example.

Sign in / Sign up

Export Citation Format

Share Document