Data Driven Calibration Approach

Paper #:
  • 2017-01-0607

Published:
  • 2017-03-28
Abstract:
Designing a control system that can robustly detect faulted emission control devices under any environmental and driving conditions is a very challenging task for any OEM. To gain confidence in the control strategy and the values of tunable parameters requires that the test vehicles are subjected to their limits during the development process. Complexity of modern powertrain systems along with the On-Board Diagnostic (OBD) monitors with multidimensional thresholds make it very difficult to anticipate all the possible worst case situations. To find the optimum solution of this problem in traditional calibration process can be very time and resource intensive. One possible solution is to take a data driven calibration approach. In this method, large amount of data is collected by collaboration of different groups working on same powertrain. Later the collected data is mined to find the optimum values of tunable parameters for respective vehicle functions. This large scale data (couple of terabytes) gives the engineers more samples to analyze and confidence in the calibration. A robust data analysis platform with the capability of handling large scale data and configurable enough to cater to different engineer’s need is required to accomplish this task. In this paper, we present some salient features of such a platform developed by Ford Motor Company. We highlight the memory efficiency, model in loop capabilities, and parallel processing capability of this platform. We also discuss the use MapReduce ideology in calibration data analysis paradigm.
Also in:
  • SAE International Journal of Commercial Vehicles - V126-2
  • SAE International Journal of Commercial Vehicles - V126-2EJ
Sector:
Access
Now
SAE MOBILUS Subscriber? You may already have access.
Buy
Attention: This item is not yet published. Pre-Order to be notified, via email, when it becomes available.
Select
Price
List
Download
$22.00
Mail
$22.00
Members save up to 36% off list price.
Share
HTML for Linking to Page
Page URL

Related Items

Technical Paper / Journal Article
2003-10-27
Article
2016-03-04
Training / Education
2010-03-15
Technical Paper / Journal Article
2004-06-08
Technical Paper / Journal Article
2004-01-16