Finite-Time Identification Algorithm for Volumetric Efficiency Map in SI Gasoline Engines

Abstract

In contrast to the calibration of an engine volumetric efficiency map (or lookup tables) from steady-state data, a method for the identification of the engine volumetric efficiency map using transient condition data is developed in this article, which can reduce the time and cost of engine map calibration. To expediently identify the map parameters, the mathematical description of the volumetric efficiency for a spark ignition (SI) gasoline engine is approximated in terms of a piecewise bilinear interpolation model with unknown map parameters, which can be rewritten as a dot product between a regression vector and a parameter vector using a membership function. With the combination of a map regression model and an SI gasoline engine air path dynamic model, a finite-time adaptive observer is designed to estimate the constant map parameters, for which the parameter estimation error information is explicitly derived. To guarantee the convergence of the identification for all map parameters, a trajectory scheme of the transient condition data from the map input that passes through all of the interpolation regions is introduced. The performance of the proposed algorithm is validated against transient condition data generated by the engine simulation software enDYNA provided by Tesis. It is shown that the volumetric efficiency map of an SI gasoline engine simulated with enDYNA is approximated acceptably by the proposed method.