Development of a Hierarchical Observer for Burned Gas Fraction in Inlet Manifold of a Turbocharged Diesel Engine

Altogether with extent of automotive role in today's life, governments are approving stringent laws to lower the permitted level of automotive emissions. An important specie of automotive engines emissions is NOx. Different methods are developed to decrease the level of NOx generation; one of which is exhaust gas recirculation (EGR). Since the performance of EGR is strictly depended on burned gas fraction (BGF) in inlet manifold, the precise control of BGF is of importance. Unfortunately, due to inlet manifold temperature, no economical sensor is available to measure the BGF; therefore, estimators are employed instead. In this paper, a stable observer is designed for estimation of air fraction in inlet manifold. The governing equations are in the form of linear parameter varying (LPV). Since the LPV parameters are not directly measured, a hierarchical estimator structure is developed. Lyapunov theory is employed to design the higher level estimator, while high gain estimators and open-loop estimators are developed for estimation of the lower level parameters. Experimental test results show that the higher level estimator is able to estimate the BGF with high accuracy in both transient and steady states. Furthermore, it is shown that BGF estimator is more sensitive to aspirated gas flow estimation rather than exhaust temperature and pressure.