Statistička analiza i modeliranje neuronskom mrežom procesa obrade komunalne vode

master's thesis

• Title (english): Statistical analysis and naural network modelling of sewer wastewater biological treatment
• Author: Anita Bevetek
• Mentor: Želimir Kurtanjek (mentor)
• Granter: University of Zagreb; Faculty of Food Technology and Biotechnology; (Department of Process Engineering); (Laboratory for Measurement, Regulation and Automatisation); Zagreb
• Defense date and country: 2005-07-08, Croatia
• Scientific / art field, discipline and subdiscipline: BIOTECHNICAL SCIENCES; Biotechnology

Abstract

Obrada otpadne vode s aktivnim muljem izrazito je složen proces, a nestalnost sastava i protoka otpadne vode te vremenski promjenjive reakcije mikroorganizama uvjetuju njegovu nelinearnost i dinamičnost. S tim u skladu i njegovo je optimalno vođenje otežano, ali pravilno izrađeni model mogao bi poslužiti kao rješenje problema. Stoga je u ovom radu izrađeno nekoliko modela za opis tog procesa na uređaju grada Velika Gorica ("Velkom"), a s ciljem predviđanja KPK vrijednosti efluenta. Izrađeni su modeli: viševarijantni linearni model sa 16 varijabli, model lokalne linearne regresije sa 17 varijabli te model RBF neuronske mreže sa 18 varijabli. Kao najuspješniji u predviđanju ponašanja procesa (korelacija 0,83) pokazao se RBF model, kojim su utvrđeni i optimalni uvjeti za vođenje procesa (protok 120 - 145 L/s i suha tvar biomase aktivnog mulja 2,5 - 4 kg/m³).

Abstract (english)

Activated sludge wastewater treatment is a highly complex process, and variations in wastewater flow rate and its composition, combined with time-varying reactions of microorganisms, make this process non-linear and dynamic. Due to this fact, process control is very sensitive but a proper model could present a solution for this problem. In this paper several models have been tested for prediction of COD in effluent of wastewater treatment plant "Velkom" in Velika Gorica. Proposed models are: Multivariate Linear Regression model (MLR) with 16 predictor variables included, Piecewise Linear Regression model (PLR) with 17 predictor variables and RBF Neural Network model (RBF NN) with 18 predictor variables included. Analysis of results show most successful prediction of process behaviour by RBF NN, with correlation 0,83, which is then used for estimating optimal operative conditions (wastewater flow rate should be in range 120-145 L/s and concentration of activated sludge dry matter between 2,5 – 4 kg/m³).

• Keywords: komunalne vode; KPK; modeliranje; neuronske mreže
• Keywords (english): COD; modelling; neural network; sewer wastewater
• Language: croatian
• URN:NBN: urn:nbn:hr:159:121423
• Study programme: Title: Biotechnology; specializations in: Biochemical Engineering; Course: Biochemical Engineering; Study programme type: university; Study level: graduate; Academic / professional title: diplomirani inženjer biotehnologije; smjer biokemijsko inženjerstvo (diplomirani inženjer biotehnologije; smjer biokemijsko inženjerstvo)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Access restricted to students and staff of home institution
• Created on: 2017-09-11 09:32:31