The open archive for STFC research publications

Full Record Details

Persistent URL http://purl.org/net/epubs/work/30124
Record Status Checked
Record Id 30124
Title The use of data mining for the monitoring and control of anaerobic waste water treatment plants
Abstract This paper describes the role of data mining in the EU-funded TELEMAC project. TELEMAC provides SMEs with telemonitoring and communications support for the deployment of anaerobic digesters for the treatment of waste water from the alcoholic beverages industry. Anaerobic digesters are very efficient but can become unstable so require expert knowledge for their handling. TELEMAC aims to enable the expert to work remotely and expertise to be shared. The purpose of data mining in this context is to provide support for the fault detection and isolation system and guidance for the experts. In this paper we discuss the role of data mining and indicate how it will be deployed within TELEMAC. We report work on identifying digester process states, sensor depletion or malfunction, and prediction intervals and prediction risk associated with neural net models. We report on using clustering and visualization to identify process states. Good estimates of chemical oxygen demand can be made using neural nets; the prediction risk and local error bars are seen to be qualitatively satisfactory.
Organisation CCLRC , BITD
Keywords Data mining , Wastewater treatment , Anaerobic digestion , Engineering , Process control
Funding Information
Related Research Object(s):
Language English (EN)
Type Details URI(s) Local file(s) Year
Paper In Conference Proceedings In Fourth European Conference on Ecological Modelling / Fourth International Workshop on Environmental Applications of Machine Learning (ECEM/EAML 2004), Bled, Slovenia, 27 Sep 2004 - 1 Oct 2004, (2004). EAMLpaper.pdf 2004
Showing record 1 of 1