State and Parameter Estimation for Spark Ignition Engine with Parameter Uncertainty

Engines are complex and their control is very important in the automotive industry. One of the main objectives while designing the control strategies is to maintain the air-fuel ratio to a specific value. This has increased the inspection of engine dynamical models and their parameter estimation. Some of the main parameters affecting the air-fuel ratio are the throttle discharge coefficient, thermal efficiency, and volumetric efficiency. These parameters are usually constant under steady-state but due to the long run of the engine and various uncertainties, their value may change. The main challenges are how to obtain the information of parameters and the states under the influence of process noise, measurement noise, and parameter uncertainty, which are required for designing an effective control strategy. In this work, the problem of physical parameter estimation of the nonlinear mean value engine system model comprising a throttle, intake manifold, engine speed dynamics, and fuel system altogether with unknown states is considered. A method with a unique combination of Unscented Kalman Filter and Recursive Least Squares with forgetting factor for estimation of parameters and states of spark-ignition engines is proposed. Simulation results are provided for state and parameter estimation for the spark ignition engine model with parameter uncertainties.