.png){kind=icon}
Return
Ca50 estimation based on Neural Network and smooth variable structure filter
.png){kind=icon}
PRE
AI
DOI：10.1016/j.isatra.2020.12.032.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
CA50 and IMEP (indicated mean effective pressure) are two critical parameters indicating the combustion process and work output. CA50 and IMEP are playing important roles in advanced combustion modeling and control for compression-ignition and spark-ignition engines. However, both CA50 and IMEP are calculated by using the in-cylinder pressure trace, the failure or aging/wearing of cylinder pressure sensors can significantly damage the accuracy of CA50 and IMEP. In this paper, the effects of different measurement error sources, especially the pressure offset and thermal shock, on CA50 and IMEP are analyzed. The results show that thermal shock reduces the IMEP calculation accuracy, while the CA50 is sensitive to all measurement errors. To improve the accuracy and reliability of CA50 calculation, a neural network was trained to capture the CA50 by utilizing other in-cylinder information, including the temperature, pressure, oxygen concentration, injection mass as well as injection time. By treating the CA50 as dummy'' state, smooth variable structure filter is used to integrate the CA50 calculated from in-cylinder pressure and the neural network. The experimental results validate the accuracy and robustness of the proposed method. (C) 2020 ISA. Published by Elsevier Ltd. All rights reserved.
Keywords:
CA50
IMEP
Neural network
SVSF
In-cylinder pressure
Journal
IF:
6.5
Papers:
6.0K
Citations:
2.0W
Organization
Cited Papers
Controlling the optimal combustion phasing in an HCCI engine based on load demand and minimum emissions.png)
ENERGY
ENERGYIF9.4
PREAI
ShareSave
ShareSaveClosed-loop combustion phase control for multiple combustion modes by multiple injections in a compression ignition engine fueled by gasoline-diesel mixtureAPPLIED ENERGY
APPLIED ENERGYIF11
PREAI
ShareSaveShareSaveShareSaveShareSave
View more