Thomas Kruse

Modern Statistical Modeling and Evolutionary Optimization Methods for the Broad Use in ECU Calibration

Abstract Model based methods can give an important contribution to cope with the drastically increased complexity in ECU calibration. This paper describes new algorithms that overcome many restrictions of classical modeling and optimization approaches. The implementation of the new algorithms in a tool framework allows a broad use in series calibration. Examples from engine base calibration (gasoline and diesel) and from drivability calibration are shown.

Calibrating Bosch Motronic systems

In 2000, Bosch and AVL began joint development of tailor-made test routine procedures for calibrating Bosch-Motronic systems. The aim was to provide worldwide support to customers calibrating their Motronic engine management systems. Bosch and AVL strove to create a solution which would provide reproducible calibration results of high quality and systemize and accelerate the calibration process. The resulting Bosch/AVL-iProcedures are available to all customers of Bosch and AVL.

Increased efficiency in the calibration process of automotive Li-ion battery systems

This paper discusses the use of model-based calibration for Li-ion batteries. The scope is set on the modeling algorithm, in this case the Gaussian Process Regression (GPR) in combination with an external dynamic structure (NARX), and its capability to describe dynamic battery behavior. A data-driven model is generated based on measurement of a Stuttgart Cycle. Model quality is evaluated on an Artemis Cycle. The achieved results show that the GPR can be used in a future model-based battery calibration process for dynamic applications in order to increase calibration efficiency.