Abstract
Three-way catalytic converters (TWC) are today's most successful aftertreatment concept for gasoline engines. Conversion rates of over 98% are reached if the exhaust gases exiting the engine are kept within a narrow band around stoichiometric conditions. However, during transient operation of the engine, which often occurs during a typical driving cycle, short deviations to the rich or to the lean side cannot be avoided by the air-to-fuel ratio (A/F) control system. As long as the oxygen storage of the catalyst is neither completely oxidized nor completely reduced, these deviations can be compensated by the three-way catalytic converter.
Thus, an additional controller is needed to keep the oxygen storage level at around half full. The oxygen storage level cannot be measured by any sensor available on the market. Because of the highly nonlinear and also asynchronous filling and depletion behavior, the controller needs to predict the actual oxygen storage level. In the past years, a deep understanding of how oxygen is stored and released in catalysts has been developed. This knowledge is now applied to design model-based TWC controllers.
A complex chemical model is simplified using only two reactions. One is the oxidation of the catalysts surface with oxygen (O2), the other its reduction with carbon monoxide (CO). The sensitivity to some of the parameters is shown. The model is verified with measurements from an engine test bench.
As a side effect, this observation of the oxygen storage level can also be used for on-board diagnosis (OBD).
Keywords: automotive, exhaust gas aftertreatment, three-way catalytic converter, control-oriented modeling