This paper presents a monitoring approach for Gasoline Particulate Filter (GPF) systems. Initially, two correlation coefficients are computed: one is between the differential pressure across the GPF and the exhaust gas volume flow rate, and the other is between the differential pressure across the muffler and the exhaust gas volume flow rate. Subsequently, the failure mode of the GPF system is identified by analyzing the distribution of these two correlation coefficients in a two-dimensional coordinate plane, where the former coefficient serves as the horizontal axis and the latter as the vertical axis. The proposed strategy was implemented in an automotive-grade controller, and on-vehicle validation tests were conducted under both the Worldwide Harmonized Light-Duty Vehicle Test Cycle (WLTC) and real-road driving conditions. The test results demonstrate that this method can accurately detect multiple failure modes, including GPF catastrophic failure, and interchange or disconnection of the GPF differential pressure sensor, thereby satisfying the requirements of worldwide OBD standards which mainly include EOBD (European On-Board Diagnostics), OBDII (On-Board Diagnostics II) and COBD (Chinese On-Board Diagnostics). Furthermore, this approach obviates the need for vehicle-level calibration of the GPF differential pressure-volume flow rate characteristic model and the muffler differential pressure-volume flow rate characteristic model. This not only reduces the consumption of resources such as prototype vehicles and chassis dynamometers but also shortens the overall development cycle time.
Deng et al. (Wed,) studied this question.