Engine cylinder pressure reconstruction using crank kinematics and recurrently-trained neural networks
- Submitting institution
-
University of Sussex
- Unit of assessment
- 12 - Engineering
- Output identifier
- 778_63580
- Type
- D - Journal article
- DOI
-
10.1016/j.ymssp.2016.07.015
- Title of journal
- Mechanical Systems and Signal Processing
- Article number
- -
- First page
- 126
- Volume
- 85
- Issue
- -
- ISSN
- 0888-3270
- Open access status
- Compliant
- Month of publication
- August
- Year of publication
- 2016
- URL
-
http://dx.doi.org/10.1016/j.ymssp.2016.07.015
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
3
- Research group(s)
-
-
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Advanced combustion control in low carbon vehicles fitted with an IC engine, requires cost-effective cylinder pressure reconstruction. A recurrent neural network is the most accurate architecture to do this but is very difficult to train recurrently. This paper configures a recurrent architecture and adapts a fully-recurrent training methodology. The paper shows that fast recurrently-trained networks are capable of accurately reconstructing the location of peak pressure (suitable for use on an engine production line). The work was funded by the EPSRC, then Jaguar Land Rover, under contract EP/E03245X/1. Contact: Mr D Richardson, Powertrain Research & Technology, Jaguar Land Rover. Email: dricha69@jaguarlandrover.com
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
