This content is not included in
your SAE MOBILUS subscription, or you are not logged in.
A Robotic Driver on Roller Dynamometer with Vehicle Performance Self Learning Algorithm
Annotation ability available
Sector:
Language:
English
Abstract
A robotic driver has been designed on the basis of an analysis of a human driver's action in following a given driving schedule. The self-learning algorithm enables the robot to learn the vehicle characteristics without human intervention. Based on learned relationships, the robotic driver can determine an appropriate accelerator position and execute other operations through sophisticated calculations using the future scheduled vehicle speed and vehicle characteristics data. Compensation is also provided to minimize vehicle speed error.
The robotic driver can reproduce the same types of exhaust emission and fuel economy data obtained with human drivers with good repeatability. It doesn't require long preparation time. Thereby making it possible to reduce experimentation work in the vehicle development process while providing good accuracy and reliability.
Authors
Topic
Citation
Moriyama, A., Murase, I., Shimozono, A., and Takeuchi, T., "A Robotic Driver on Roller Dynamometer with Vehicle Performance Self Learning Algorithm," SAE Technical Paper 910036, 1991, https://doi.org/10.4271/910036.Also In
References
- Gryce R.T. “Ford Auto/Emission Driver System,” SAE 741007
- Willn J.E. Forbes R.T. Oberoi S.N. Patullo J.S. “A Robotic Driver Consumption and Drive-line Testing,” SAE 851640
- Valette R. Gasnier M. “An Automatic Drive System for Chassis Dynamometer Vehicle Testing,”
- Federal Register 12 124 June 26 1977
- Fock Martin Lies Karl-Heinz Pazsitka Laszlo “Critical Study of the United States Exhaust Emission Certification Test-Error and Probability Analysis,” SAE 750678
- Juneja Wiplove K. Horchler David D. Haskew Harold M. “A Treatise on Exhaust Emission Test Variability,” SAE 770136