Investigation of a Model-Based Approach to Estimating Soot Loading Amount in Catalyzed Diesel Particulate Filters

In order to meet the worldwide increasingly stringent particulate matter (PM) and particulate number (PN) emission limits, the diesel particulate filter (DPF) is widely used today and has been considered to be an indispensable feature of modern diesel engines. To estimate the soot loading amount in the DPF accurately and in real-time is a key function of realizing systematic and efficient applications of diesel engines, as starting the thermal regeneration of DPF too early or too late will lead to either fuel economy penalty or system reliability issues. In this work, an open-loop and on-line approach to estimating the DPF soot loading on the basis of soot mass balance is developed and experimentally investigated, through establishing and combining prediction models of the NO_x and soot emissions out of the engine and a model of the catalytic soot oxidation characteristics of passive regeneration in the DPF. The emission testing results under the New European Driving Cycle (NEDC) show that the prediction errors of the engine-out NO_x and soot emission models are 5.1% and 3.9%, respectively. Tests and validations of the soot mass loading model are carried out under on-vehicle driving. The experimental results show that the maximum estimation error of the model is 0.48g/L and the average error is 0.17g/L. It shows that the model estimation error is less than 6%, which is conducive to promoting safe and reliable DPF regeneration and contributes to the DPF management and applications in real-world operation.