Feasibility study of the potential use of chemistry based emission predictions for real-time control of modern diesel engines

The feasibility of using chemical kinetics-based prediction of emission species for real-time control of modern diesel engines is investigated. A previously developed fast, physics-based model is used as a representative example. The temporal variation of temperature required for the computation of the reaction rate constants is obtained from the solution of the energy equation. The effects of composition and temperature on the thermo-physical properties of the working fluid are included in the computations. Issues relating to model complexity, computation time, and fidelity are discussed in the context of both equilibrium and finite rate chemistry for use in the real time environment. The set of model inputs and tunable parameters is assessed for real-time use against the standard sensor set available on modern diesel engines. Results show that use of physics-based quasi-dimensional models is promising but may need complex variable mappings for real-time application.

► We discuss the essential trade-offs in the selection of the best modeling approach for RTC.
► We examine a kinetics-based model in the context of RTC.
► We showed that physics-based models need a reduced tunable parameter set for RTC.
► We found that use of physics-based models can greatly reduce the calibration burden for RTC.
► We found that use of physics-based quasi-dimensional models may need complex variable mapping.