Search

• A method for trapped mass estimation from in-cylinder pressure resonance is used. • An iterative algorithm is developed for reducing the computational burden. • An outlier rejection procedure is presented for improving the robustness. • An observer is designed to improve the estimation in a multi-sensor scenario. A novel technique of trapped mass determination, based on the in-cylinder pressure resonance, has been recently published by the authors. However, the method only works when sufficient resonance intensity exists and the current formulation might preclude its implementation in real-time due to excessive computational burden. The present paper proposes an iterative algorithm for reducing the number of operations, an adaptive filter to identify faulty measurements and a Kalman filter that combines several sensors and models, currently used in commercial light-duty engines, to ensure a continous estimation of trapped mass, air mass, and exhaust gas recirculation (EGR). The filter is implemented using experimental data of a EURO 6 light-duty engine in a world harmonize light-duty test cycle (WLTC), showing the potential of being implemented in real driving conditions with robustness and harnessing a new measurement to improve the accuracy and response of current estimations. [ABSTRACT FROM AUTHOR]