Project sponsored by Eminox

Suitable for students with Engineering, Chemistry or Mathematical Sciences backgrounds

Rapid Modelling and Calibration of Aftertreatment Systems

Modern Diesel engines use a number of catalysts and filters to reduce harmful emissions. A recent trend is to combine the Selective Catalytic Reduction (SCR) with the Diesel Particulate Filter (DPF) in one SCR-DPF brick. While highly effective, such a system is inherently difficult and time consuming to model. This project will investigate ways of formalising and automating the modelling process in order to reduce the calibration time. Possible approaches include automatic experiment design, dynamic model fitting, self-learning models, and adaptive control strategies.

Background

Heavy duty internal combustion engines are tuned for high efficiency at typical operating points. A side effect of this is that significant amounts of toxic emissions are generated, which requires careful management of the aftertreatment system.

Selective catalytic reduction (SCR) is an approach commonly applied to control NOx emissions from heavy duty diesel engines. The SCR catalyst can also be applied to Diesel particulate filters (DPF) to provide an integrated aftertreatment system (SCR-DPF).

The design and control of these of heavy duty SCR-DPF installations can be highly application specific, because every work load has different effects on the exhaust temperature and composition.

Aims and Objectives

There is a huge opportunity for improving NOx conversion and reducing size and weight of the SCR-DPF aftertreatment system if gas phase parameters can be controlled dynamically. This requires dedicated sensors and actuators to measure and influence gas phase parameters in order to create optimal conditions with the given physical limits.

The main research questions to consider are:

• How detailed does the model need to be for effective control?
• How can the model parameters be found in an efficient manner?
• Can adaptive control help with the modelling and with the long term control?
• Which structures and sensors need to be in place for the system to monitor itself?

Work Programme

The work programme will be updated throughout the four year project, but it is expected to include the following aspects:

1. Compare phenomenological and physical models for SCR-DPF aftertreatment systems, focussing on complexity and accuracy. Identify relevant test data.
2. Consider a number of control strategies, which provide both adaptation and a degree of robustness. The information problem may require additional sensors or additional excitation, which may be contrary to the control goals. This conflict needs to be explored.
3. Create a modular model and a control library for easy adaptation to different circumstances.
4. Analyse the impact of the control strategies from a number of perspectives, including performance, reliability, robustness and commercial viability. Communicate the results for further commercialisation, and disseminate the scientific findings.

The Engineering Doctorate (EngD) is of four years duration, and carries an enhanced annual stipend of £18,363.

To apply send your CV and covering letter to the EPSRC Centre for Doctoral Training in Carbon Capture and Storage and Cleaner Fossil Energy: ccscfe<στο>nottingham.ac.uk

Please quote ref: ENG830                             Closing date: 8 April 2015