Artificial intelligence for the ageing identification of exhaust gas catalytic converters as part of the KIARA research project

As part of the KIARA research project funded by the Federal Ministry of Economics and Climate Protection (BMWK): "Artificial intelligence for ageing identification of real applications of exhaust gas catalytic converters", IVD Deutschland GmbH is working together with the Institute for Internal Combustion Engines and Vehicle Drives (VKM) to develop an AI-based method for analyzing and reducing ageing phenomena in the exhaust gas aftertreatment system.

The increasing pressure on politicians and society to shape the path to sustainable mobility can only be realised to a limited extent through electromobility. In a CO₂ balancing in the sense of a complete well-to-wheel consideration, purely electrically powered vehicles show disadvantages compared to conventional forms of propulsion based on the use of liquid fuels. These disadvantages lie in energy generation and distribution, which must, however, be taken into account from a climate impact perspective. Drive systems of future vehicles in the passenger car and commercial vehicle sectors will therefore continue to have a high proportion of internal combustion engines. Today's vehicles with modern emission reduction systems already offer a very low level of emissions.

As a result of increasing vehicle mileage, the high degree of pollutant conversion in the catalytic converters of the exhaust gas purification system can decrease depending on the thermal, chemical and mechanical load. This effect, known as "ageing", cannot yet be adequately recorded due to the high variance in usage profiles caused by vehicle functions. For this reason, legal requirements for so-called "in-service conformity" (ISC), which are already being discussed, pose new challenges for the development of sustainable exhaust gas aftertreatment systems. Catalytic converter and filter systems must be stable in their performance in terms of pollutant conversion even after long periods of use, many vehicle kilometres and, above all, very unpredictable operating conditions in order to make a major contribution to sustainability.

In addition to recognising ageing patterns of the individual components of the exhaust gas purification system during driving, the AI-based method is intended to help reduce the ageing of the exhaust gas aftertreatment system through targeted interventions in the operating strategy of the engines. For an optimum sustainability profile, it is crucial to minimise exhaust emissions and achieve a high level of catalytic efficiency even in an aged state.

In addition to the road test with the highly dynamic 4×4 roller test bench at the Technology and Environmental Centre in Pfungstadt (TUZ GmbH) and an engine-in-the-loop engine test bench at VKM, an optimal development environment with extensive exhaust gas measurement technology is available for training and continuous further development of AI-based ageing detection.

In addition to the ageing states measured on the engine-in-the-loop engine test bench and the real road journeys, reproducible cycles in the real vehicle can now be used to identify ageing. The degree of repeatability can be further increased by using a driving robot. The tests on the chassis dynamometer, with their clear definition and high repeat accuracy, can be used to precisely determine the condition of the exhaust aftertreatment systems that have aged in the field and thus quantify and classify ageing. The ageing processes calculated in parallel on the basis of the AI model can be compared with the measured data and the AI-based model can be further refined and optimised with regard to its quality.

The aim of AI-based ageing detection is to guarantee the lowest exhaust emissions even when the exhaust gas aftertreatment system is aged by making targeted interventions in the operating strategies of the engines. This serves to differentiate and set the system apart from ageing detection systems currently used in the field, which allow purely static interventions in the operating strategies of the engines but have no relation to emissions. 

Contact  

Institut für Verbrennungskraftmaschinen und Fahrzeugantriebe

Technische Universität Darmstadt

Otto-Berndt-Straße 2

64287 Darmstadt

Mr. Herr M.Sc. Luis Vincent Fiore  

Tel. +49 6151 16 21376

fiore@vkm.tu-darmstadt.de