Control Oriented Physics Based Three-Way Catalytic Converter Temperature Estimation Model for Real Time Controllers 2020-01-0904
As automotive emissions become more stringent, accurate control of three-way catalyst temperature is increasingly important for maintaining high levels of conversion efficiency as well as preventing damage to the catalyst. A real-time catalyst temperature model provides critical information to the engine control system. In order to improve emissions and ensure regulatory compliance over a wide range of speed-load conditions, it is desirable to use modelled catalyst temperature as the primary input to catalyst efficiency control strategies. This requirement creates a challenge for traditional empirical models designed for component protection at high speed-load conditions. Simulation results show that a physics aligned model can estimate temperature in all operating conditions, including: cold-start, extended idle, engine shutdown, stop-start events, decel fuel shut-off, as well as traditional high load and part load points. However, physics based approaches which calculate detailed chemical reaction kinetics remain impractical for real-time controller implementation due to computational burden and calibration complexity. This paper outlines a proposal for a simplified control-oriented physics model which estimates catalyst temperature in real time. The model consists of reduced order exothermic reaction kinetics in conjunction with gas phase and solid phase energy balance equations. The control-oriented physics model is verified in both steady-state and transient drive cycles and shown to provide increased estimation accuracy while reducing transient calibration effort when compared to a traditional empirical model.
Darshan Shah, Kiran Premchand, David Pedro