Experimental Design and Development of an Expert System for the Knowledge-Based Engine Process Optimization of Modern Diesel Engines

This paper presents a new approach to the optimization process of modern diesel engines. The establishing of an automated online calibration demands an accurate optimization method with the ability to handle the increasing complexity of diesel engine concepts and to minimize the calibration time by rapid calculations. Model-based optimziation algorithms used for offline optimization so far do not come up to these requirements. Starting from an advanced design of experiment a knowledge-based expert system has been developed for emulating the decision-making process and the optimization method of an application engineer. The knowledge-based method is exemplarily applied to a stationary operating point.