Xing, Yuxin (2019) Micro-hotplate based CMOS sensor for smart gas and odour detection. PhD thesis, University of Warwick.

Preview

PDF WRAP_Theses_Xing_2019.pdf - Submitted Version - Requires a PDF viewer. Download (23Mb) | Preview

Request Changes to record.

Abstract

Low cost, highly sensitive, miniature CMOS micro-hotplate based gas sensors have received great attention recently. The global sensor market is expanding rapidly with an expected increase of 5 ~ 8% grow thin the next five years. The application areas for a gas sensor include but are not limited to, air quality monitoring, industrial and laboratory conditions, military, and biomedical sectors. It is the key hardware component of an electronic nose, as well as the signal processing on the software side. In this thesis, both aspects of such a system were studied with new sensor technologies and improved signal processing algorithms. In addition, this thesis also described different applications and research projects using these sensor technologies and algorithms.

A novel plasmonic structure was employed as an infrared source for anon- dispersive infrared gas sensor. This structure was based on a CMOS micro hot plate with three metal layers and periodic cylindrical dots to induce plasmon resonance, that allowed a tunable narrow band infrared radiation with high sensitivity and selectivity. Five gases were studied as target gases, namely, carbon monoxide, carbon dioxide, acetone, ammonia and hydrogen sulfide. These emitter sources were fabricated and characterised with a gascell, optical filters and commercial detectors under different gas concentrations and humidity levels. The results were promising with the lowest detection limit for ammonia at 10 ppm with 5 ppm resolution. On the data processing side, various signal processing methods were explored both on-board and on-board. Temperature modulation was the on-board method by switching the operating temperatures of a micro hotplate. This technique was proven to over come and reduce some typical sensor issues, such as drift, slow re-sponse/recovery speed (from tens of seconds to a few seconds) and even cross sensitivities. Off-board post processing methods were also studied, including principal component analysis, k-nearest neighbours, self-organising maps and shallow/deep neural networks. The results from these algorithms were compared and overall an 85% or higher classification accuracy could be achieved. This work showed the potential to discriminate gases/odours, which could lead to the development of a real-time discrimination algorithm for low cost wearable devices.

Item Type:	Thesis or Dissertation (PhD)
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering T Technology > TP Chemical technology
Library of Congress Subject Headings (LCSH):	Gas detectors -- Design and construction, Olfactory sensors -- Design and construction, Metal oxide semiconductors, Complementary, Odors, Volatile organic compounds, Mobile robots -- Remote sensors
Official Date:	December 2019
Dates:	Date Event December 2019 UNSPECIFIED
Institution:	University of Warwick
Theses Department:	School of Engineering
Thesis Type:	PhD
Publication Status:	Unpublished
Supervisor(s)/Advisor:	Gardner, Julian ; Cole, Marina
Sponsors:	University of Warwick. School of Engineering ; Horizon 2020 (Programme) ; COMET IC-MPPE (Programme)
Format of File:	pdf
Extent:	xxxii, 249 leaves : illustrations
Language:	eng

Request changes or add full text files to a record

Repository staff actions (login required)

View Item

Downloads