Modeling and Observability Analysis of DPF Regeneration

A Diesel Particulate Filter (DPF) is a necessary component on all modern Diesel vehicles in order to achieve the emission targets for particulate matter. A key problem in this context is the accurate real-time estimation of the soot in the DPF during or after a regeneration event. In this paper, a lumped parameter model for the regeneration regime of a Diesel Particulate Filter (DPF) is proposed which can be used as a starting point for the above purpose. Sensitivity of this model to various model parameters is discussed and significant parameters are identified. An observation problem is identified and a preliminary observability analysis is performed. Results of model performance and implications of the observability analysis are presented.