Dp 579 en
Modeling and state estimation of automotive SCR catalyst
Author: Figura Jiří
The amount of NH3 stored on the walls of Selective Catalytic Reduction (SCR) systems is an interesting parameter for utilization in urea injection control. However, it has to be estimated as it cannot be measured onboard vehicle.This thesis introduces the state of the art in modeling and estimation of SCR systems by reviewing the published literature. Further, it proposes a nonlinear physics-based SCR catalyst model and Extended Kalman Filter (EKF) for online estimation of outlet specie concentrations and NH3 storage.
Simulations over standardized driving cycle show the designed model provides accurate openloop predictions and the EKF further slightly improves them. A robustness analysis shows that, despite sensitivity to inputs, EKF introduces significant benefit over open-loop model simulations.
- Jiří Figura, mailto:figurjir@fel.cvut.cz
