Title :
LSSVM based dynamic control model of activated sludge process
Author :
Yuling Tang ; Xiaoxuan Zhang ; Shirong Zhang
Author_Institution :
Dept. of Autom., Wuhan Univ., Wuhan, China
Abstract :
Activated sludge is a typical process for waste water treatment; meanwhile, it is an energy-intensive process. Operation optimization is a solution to save energy, where control models are much needed. In this paper, a dynamic LSSVM model for activated sludge process is proposed and validated. Time delay and model orders are involved during the re-establishment process of the training datasets; by this way, the dynamic characteristic of the process is reflected. Finally, BSM1 will be taken as the objective process for validation research, along with convincing results.
Keywords :
environmental science computing; least squares approximations; sludge treatment; support vector machines; wastewater treatment; BSM1; LSSVM based dynamic control model; activated sludge process; energy-intensive process; least squares support vector machine; model orders; operation optimization; process dynamic characteristic; time delay; training datasets; waste water treatment; BSM1; Dynamic; LSSVM; Performance; Waste Water Treatment;
Conference_Titel :
Fluid Machinery and Fluid Engineering, 2014 ISFMFE - 6th International Symposium on
Print_ISBN :
978-1-84919-907-0
DOI :
10.1049/cp.2014.1242
