Anomaly detection of municipal wastewater treatment plant operation using support vector machine

Z. X. Tian, J. P. Jiang, L. Guo, P. Wang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

It is difficult to run a wastewater treatment process (WWTP) stably in the long term. In this work to monitor the operation state of the treatment process Support Vector Machine is applied for anomaly detection in Municipal Wastewater Treatment Plant based on operational data. Considering the characteristics of the water quality parameters and relevant regulations, we select the detection vector and choose C-SVM and Radial Basis Function (RBF).Then this paper analysis the parameters optimization of SVM, using the Grid search and Particle Swarm Optimization for model calibration. By comparing the accuracy with 10-Cross Validation of three models, we determine the final classification model. Validation demonstrates the model is able to gain high classification accuracy.

Original languageEnglish
Title of host publicationInternational Conference on Automatic Control and Artificial Intelligence, ACAI 2012
Pages518-521
Number of pages4
Edition598 CP
DOIs
StatePublished - 2012
Externally publishedYes
EventInternational Conference on Automatic Control and Artificial Intelligence, ACAI 2012 - Xiamen, China
Duration: 3 Mar 20125 Mar 2012

Publication series

NameIET Conference Publications
Number598 CP
Volume2012

Conference

ConferenceInternational Conference on Automatic Control and Artificial Intelligence, ACAI 2012
Country/TerritoryChina
CityXiamen
Period3/03/125/03/12

Keywords

  • Anomaly Detection
  • Municipal Wastewater Treatment Plant (MWWTP)
  • Parameters Optimization
  • SVM

Fingerprint

Dive into the research topics of 'Anomaly detection of municipal wastewater treatment plant operation using support vector machine'. Together they form a unique fingerprint.

Cite this