A temporal discretization and spatial integration SCR model with dual temperature-related parameters

Abstract

Developing an accurate and rapid Selective Catalytic Reduction (SCR) model is crucial for advanced control strategies, which is challenging due to the complex physicochemical processes. In this paper, a novel control-oriented SCR model is proposed, which exhibits faster and more accurate estimation performance for both NOx and NH3 emissions. Unlike traditional modeling methods, this research develops a temporal discretization and spatial integration (TDSI) SCR model with dual temperature-related parameters. By introducing NH3 recycling, the spatial integration model is improved, leading to an enhanced accuracy in model estimation. Moreover, A time discretization method based on current state can effectively address the issue of sampling time sensitivity. Additionally, dual temperature-related parameters can significantly enhance model accuracy, particularly in high-temperature regions. A simplified and clear flowchart is provided for parameter identification based on a limited number of engine bench test results. In a cold-start World Harmonized Transient Cycle (WHTC), the TDSI model can achieve NOx and NH3 accuracies of 92.26% and 92.52%, respectively, with a computation time of 1.47 s. Through comprehensive comparisons, the proposed model exhibits comparable accuracy to the GT model while significantly reducing computation time by 98.39%. Additionally, compared to the Continuous Stirred-Tank Reactor (CSTR) and θT models, there is an approximate 3% improvement in NOx and NH3 accuracies. This TDSI model provides a foundation for designing urea injection control strategies and offers new perspectives for simplifying the model design.