This content is not included in
your SAE MOBILUS subscription, or you are not logged in.
Automated Data Screening for Steady-State Engine Mapping
Annotation ability available
Sector:
Language:
English
Abstract
Software has been developed to carry out data screening on steady-state engine dynamometer data, for application to gasoline engine mapping. The software carries out fully automatic data screening, capturing the expertise of experienced engineers, with benefits in labour saving, faster throughput and improved quality of result. The technical basis of the screening process is the local regression of test results, to ensure that the engine is displaying expected physical patterns of behaviour. Emissions and combustion pressure data are screened, as well as standard parameters such as fuel flow and torque. Automated screening makes it possible to cope with much faster rates of data acquisition without needing extra personnel, and without allowing the quality of result to be compromised. The programs are currently being used on a VAX mainframe, but they are also being incorporated into a PC-based system. This will be used in dynamometer control rooms, at several test sites, in order to highlight faults immediately, and to request repeat testing before the test conditions are changed. The screening process also provides an objective measure of “goodness of fit” for the multivariate polynomial regression which is used at the end of the mapping process. This is achieved by using local regression to make objective comparisons between raw and regressed data. As new engine models are developed, they can be validated against existing raw data by the same technique.
Recommended Content
Authors
Topic
Citation
De Salis, R., "Automated Data Screening for Steady-State Engine Mapping," SAE Technical Paper 930394, 1993, https://doi.org/10.4271/930394.Also In
References
- Cassidy, John F GM Research Labs A computerized on-line approach to calculating optimum engine calibrations SAE 770078
- Mencik, Z Blumberg, P.N. Ford E&RS Representation of engine data by multi-variate least-squares regression SAE 780288