Improving the NO x reduction performance of an Euro VI d SCR System in real-world condition via nonlinear model predictive control

Abstract

This paper deals with the possibility of improving the urea dosage control for the Selective Catalytic Reduction Systems (SCR) of an Euro VI d diesel light commercial vehicle in order to increase $N{O}_x$ after-treatment reduction performance. To this aim, first, we assess the effective emissions abatement performance for the appraised diesel vehicle via real-world experimental campaign, carried out according to the Real Driving Emissions (RDE) tests on urban, extra-urban and motorway road sections in Naples, Italy. Based on these real-world data, we derive a parameterized control-oriented model for the SCR system which is, then, exploited for the designing of an alternative urea injection logic which could be able to maximize the $N{O}_x$ reduction efficiency while minimizing tailpipe ammonia slip. Specifically, the optimal AdBlue injection rate is designed according to a Nonlinear Model Predictive Control Approach which allows obtaining a proper trade-off between the $N{O}_x$ abatement and the urea overdosing problem. The effectiveness of the proposed controller is evaluated by comparing the performance assessed for the appraised SCR system during the experimental tests with the ones achievable if the Euro VI diesel would be equipped with the proposed control strategy. Numerical simulation discloses the effectiveness of the NMPC controller in ensuring improved $N{O}_x$ reduction with performance complying with the emissions norms, main in avoiding excessive ammonia slip and in guaranteeing a reduced feed ratio w.r.t. to the standard industrial SCR controller mounted on the vehicle.