Fuzzy pattern recognition of circadian cycles in ecosystems

Simple item page
dc.contributor.authorMarsili-Libelli, S.en
dc.contributor.departmentSustainable Agriculture and Natural Resource Management (SANREM) Knowledgebaseen
dc.date.accessioned2016-04-19T18:55:30Zen
dc.date.available2016-04-19T18:55:30Zen
dc.date.issued2004en
dc.description.abstractMany ecological variables show a wide range of fluctuations, the most important of which is the diurnal variation. This cycling may contain important information regarding the ecosystem's functioning and, if property interpreted, can represent a valuable predictive tool in ecosystems management. This paper describes a simple algorithm for extracting meaningful information from daily cycles using fuzzy pattern recognition techniques. The algorithm is organised in three parts: in the first, typical patterns are extracted from experimental data to form the knowledge-base upon which the algorithm operates. The second step is to condense the information contained in the knowledge-base into mathematical objects, referred to in the paper as fuzzy masks. The third step is the Set-Lip of an inferential set of fuzzy rules, using the fuzzy masks as antecedents. Depending on how the inference engine is structured, the algorithm Output can be viewed as an assessment of the current daily cycle with respect to a set of "typical" system behaviours and decisions can be made accordingly. Two applications are presented to demonstrate the algorithm. In the first a medium-scale wastewater treatment plant with minimum instrumentation is considered. Based on diurnal flow variations of the previous two days, the algorithm determines the current treatment level in terms of sludge recycling. The other application is the prediction of the eutrophication level in a coastal lagoon. In this case, the algorithm considers the diurnal fluctuations of basic water quality parameters such as dissolved oxygen, oxidation-reduction potential, pH and temperature to detect the possibility of a macroalgae bloom. In both cases. the algrithm provides information about the shift of the system from one trend to another on the basis of the analysis of diurnal cycles. The algorithm performance is assessed against practical situations and the results discussed.en
dc.format.mimetypeapplication/pdfen
dc.identifier999en
dc.identifier.citationEcological Modelling 174(1-2): 67-84en
dc.identifier.doihttps://doi.org/10.1016/j.ecolmodel.2003.12.043en
dc.identifier.issn0304-3800en
dc.identifier.other999_00011.pdfen
dc.identifier.urihttp://hdl.handle.net/10919/65878en
dc.language.isoen_USen
dc.publisherAmsterdam, Netherlands: Elsevier Science B.V.en
dc.rightsIn Copyrighten
dc.rights.holderCopyright 2004 by Elsevier B.V. All rights reserved.en
dc.rights.urihttp://rightsstatements.org/vocab/InC/1.0/en
dc.subjectSocial impactsen
dc.subjectEcosystemen
dc.subjectAquatic ecosystemsen
dc.subjectEnvironmental impactsen
dc.subjectEnvironmental servicesen
dc.subjectModelingen
dc.subjectHealth impactsen
dc.subjectResource management toolsen
dc.subjectNatural resource managementen
dc.subjectDiurnal variationen
dc.subjectPattern recognitionen
dc.subjectDecision support systems (DSS)en
dc.subjectFuzzy systemsen
dc.subjectEutrophicationen
dc.subjectPhytoplanktonen
dc.subjectWastewater treatmenten
dc.subjectLagoonsen
dc.subjectEcosystemen
dc.titleFuzzy pattern recognition of circadian cycles in ecosystemsen
dc.typeArticle - Refereeden
dc.type.dcmitypeTexten

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
999_00011.pdf
Size:
464.67 KB
Format:
Adobe Portable Document Format
Download

Collections

Sustainable Agriculture and Natural Resource Management (SANREM) Knowledgebase