Virtual Sensors to Reduce Cost for Turbo Charged Gasoline Engines
Track: Simcenter User Connection
Modern engines are equipped with many sensors to achieve a good balance of fuel efficiency, emissions, and performance. On the other hand, sensors are expensive and replacing physical sensors with virtual sensors is an intriguing option. This paper presents a practical approach to implement a virtual exhaust pressure sensor for a turbo-charged GDI engine using model-based observers. A key enabler to developing virtual sensors is using Amesim to quickly and accurately describe engine physics. This approach allows engineers to develop an Extended Kalman Filter which uses available measurements and an embedded non-linear model to estimate exhaust pressure. The virtual sensor is validated with experimental data and further work is focusing on real-time deployment on an ECU.