Exhaust Temperature Predictor: an alternative approach to predictive control applied to Diesel aftertreatment

摘要

To meet emission legislations on Diesel vehicles, Nitrogen Oxides (NO_x) aftertreatment becomes more and more important. In the configuration described, a Selective Catalytic Reduction (SCR) catalyst is located downstream a Diesel Oxidation Catalyst and a Diesel Particulate Filter. These high thermal inertia components have to be heated up before the SCR can light-off and achieve the required conversion efficiency. A discrete mode control approach is used, consisting of multi combustion modes dedicated to specific heat-up measures. The optimum compromise between heat-up time and fuel economy is then achieved by automatic switching between heat-up and normal combustion modes. A prediction of the SCR temperature is used to determine when to switch from one mode to the other. This method embeds a simplified thermal model based on the physics of the exhaust line, running hundred times faster than real time in order to achieve a prediction horizon of several hundreds of seconds. This predictor technique was applied to a FTP-75 test cycle and proved robustness on different driving cycles.