Virtual Sensors in Small Engines – Previous Successes and Promising Future Use Cases
- Delivery
- Available on this site
- Format
- Price
- Non-members (tax incl.)：￥1,100 Members (tax incl.)：￥880
- Publication code
- 20239537
- Paper/Info type
- SETC
No.2023-01-1837
- Pages
- 1-10(Total 10 p)
- Date of publication
- Oct 2023
- Publisher
- JSAE & SAE
- Language
- English
- Event
- Small Powertrains and Energy Systems Technology Conference 2023
Detailed Information
| Author(E) | 1) Andreas Benjamin Ofner, 2) Jonas Sjoblom, 3) Stefan Posch, 4) Markus Neumayer, 5) Bernhard Geiger, 6) Stephan Schmidt |
|---|---|
| Affiliation(E) | 1) Know-Center GmbH, 2) Chalmers Univ of Technology, 3) LEC Gmbh, 4) Graz University of Technology, 5) Know-Center GmbH, 6) Graz University of Technology |
| Abstract(E) | Virtual sensing, i.e., the method of estimating quantities of interest indirectly via measurements of other quantities, has received a lot of attention in various fields: Virtual sensors have successfully been deployed in intelligent building systems, the process industry, water quality control, and combustion process monitoring. In most of these scenarios, measuring the quantities of interest is either impossible or difficult, or requires extensive modifications of the equipment under consideration - which in turn is associated with additional costs. At the same time, comprehensive data about equipment operation is collected by ever increasing deployment of inexpensive sensors that measure easily accessible quantities. Using this data to infer values of quantities which themselves are impossible to measure - i.e., virtual sensing - enables monitoring and control applications that would not be possible otherwise. In this concept paper, we provide a short overview of virtual sensing and its applications in engine settings. After reviewing the current state-of-the-art, we introduce several virtual sensor use cases that have successfully been deployed in the past. Starting from a simple phenomenological model connecting the ion current from a spark plug with fuel quality, we move over physical models that infer in-cylinder pressure from the acceleration signal of knock sensors to a deep learning model that estimates combustion parameters from the vibration of the crank shaft. In this manner, this study is designed as a “teaser”, with the intention of incentivizing further development within the sector by providing the aforementioned information. We close the paper by discussing possible applications of virtual sensing in small engines. |
