DocumentCode
3057313
Title
The Study of Compost Quality Evaluation Modeling Method Based on Classify Support Vector Machine for Sewage Treatment
Author
Tian, Jingwen ; Feng, Qian ; Gao, Meijuan
Author_Institution
Dept. of Autom. Control, Beijing Union Univ., Beijing
fYear
2007
fDate
14-17 Sept. 2007
Firstpage
111
Lastpage
116
Abstract
Because of the complicated interaction of the sludge compost components, it makes the compost quality evaluation system appear the characters of non-linearity, randomness and uncertainty. According to the physical circumstances of sludge compost, a compost quality evaluation modeling method based on support vector machine (SVM) is presented. We select the index of sludge compost quality and take the high temperature duration, degrdn rate, nitrogen content, average oxygen concentration and maturity degree as the evaluation parameters. We construct the structure of SVM network that used for the quality evaluation of sludge compost and use the genetic algorithm (GA) to optimize SVM parameters. With the ability of strong self-learning and well generalization of SVM, the modeling method can truly evaluate the sludge compost quality by learning the index information of sludge compost quality. The experimental results show that this method is feasible and effective.
Keywords
genetic algorithms; sewage treatment; sludge treatment; support vector machines; unsupervised learning; compost quality evaluation modeling; degradation rate; genetic algorithm; high temperature duration; optimisation; self-learning; sewage treatment; sludge; support vector machine; Artificial neural networks; Environmental economics; Genetic algorithms; Nitrogen; Quality management; Sewage treatment; Support vector machine classification; Support vector machines; Temperature; Uncertainty;
fLanguage
English
Publisher
ieee
Conference_Titel
Bio-Inspired Computing: Theories and Applications, 2007. BIC-TA 2007. Second International Conference on
Conference_Location
Zhengzhou
Print_ISBN
978-1-4244-4105-1
Electronic_ISBN
978-1-4244-4106-8
Type
conf
DOI
10.1109/BICTA.2007.4806430
Filename
4806430
Link To Document