Advertisement

Online-methods for engine test bed measurements considering engine limits

  • Benjamin Hartmann
  • P. Heuser
  • E. Kloppenburg
  • R. Diener
Conference paper
Part of the Proceedings book series (PROCEE)

Abstract

In this contribution a method called Online-DoE with Constraint Modeling (ODCM) for online experimental design is presented. The algorithm calculates a probability value during the measurement, if planned measurement points will be drivable or nondrivable. Therefore, points that are rated as non-drivable with a high probability will be skipped.

The key innovation of the proposed approach lies in the online identification of a nonlinear global classification model which separates the drivable from the non-drivable input regions. As discussed in this contribution, the working principle of the proposed method can be distinguished from other online methodologies like Online Boundary Identification, Active Learning or Online Optimization.

The Online-DoE with Constraint Modeling (ODCM) method has been established for several calibration projects at Bosch Engineering GmbH. The main goal of this algorithm is to improve the efficiency of measurements with space-filling DoE plans that are required for modeling and optimization. The functioning and the benefits of the proposed method are shown on simulations as well as with experimental results from engine test bed measurements.

The following advantages of the ODCM method can be summarized:

– Reduced measurement times.

– Reduced effort for planning the experiment.

– Good coverage of the drivability space.

– Reduced risk for engine stoppage during measurement.

– Better exploitation of test bed resources.

Keywords

Robert Bosch GmbH Engine Limit Calibration Engineer Sobol Sequence Gaussian Process Regression Model 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer Fachmedien Wiesbaden 2016

Authors and Affiliations

  • Benjamin Hartmann
    • 1
  • P. Heuser
    • 1
  • E. Kloppenburg
    • 2
  • R. Diener
    • 2
  1. 1.Bosch Engineering GmbHAbstattGermany
  2. 2.Robert Bosch GmbHStuttgartGermany

Personalised recommendations