Identification of the relevant parameters of the wall-wetting system by extended Kalman filtering

• Published in: Control engineering practice Vol. 14; no. 3; pp. 235 - 241
• Main Authors: Locatelli, Marzio, Alfieri, Ezio, Onder, Christopher H., Geering, Hans P.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 2006
• Subjects: Intelligent knowledge-based systems; Kalman filters; Model-based identification; Wall-wetting dynamics; Intelligent knowledge-based systems; Kalman filters; Wall-wetting dynamics; Model-based identification
• ISSN: 0967-0661; 1873-6939
• DOI: 10.1016/j.conengprac.2005.03.014

The knowledge of the wall-wetting system in an SI engine is important for a correct dosage of the fuel to be injected. The existing models are divided into two categories. The control-oriented models have parameters which are not constant over the whole operating range but are simple to parametrize. The models based on physics have a good precision, but they are not suitable for controller design. To combine the positive aspects of both model types, a physical model of the wall-wetting system whose parameters can be adjusted on-line with the aid of extended Kalman filtering was developed. This paper shows a way to implement a Kalman filter combined with an off-line identification to estimate the wall-wetting parameters and the wall film mass on the basis of this model.