Observer design for exhaust gas recirculation rate estimation in a variable-geometry turbocharger diesel engine using a model reference identification scheme

Abstract

Exhaust gas recirculation systems are used in diesel engines to reduce nitrogen oxide emissions. Since excessive recirculation in the cylinders may lead to an increase in generation of particulate matter and to unstable combustion, the exhaust gas recirculation rate should be measured correctly and should be controlled precisely. Unfortunately, the harsh conditions of the exhaust gas recirculation path make it difficult to measure the exhaust gas recirculation mass flow rate directly by using the relevant sensors. Therefore, precise control of the exhaust gas recirculation system depends on accurate estimation of the exhaust gas recirculation rate. To estimate accurately the exhaust gas recirculation rate in a turbocharged diesel engine, we propose an observer based on a model reference identification scheme. A linear parameter-varying model of the intake manifold pressure was derived to serve as the observer’s reference model. An update rule of the observer was designed with the model reference identification scheme. The intake and exhaust temperature models were developed through an empirical approach. Convergence of the proposed observer was proven in terms of the Lyapunov stability criterion. The proposed observer was implemented on a real-time embedded system and validated successfully through experiments on the engine.