Entity

Time filter

Source Type

İzmit, Turkey

Koni M.,Takosan A.S. | Yuzgec U.,Bilecik University | Yuzgec U.,Kocaeli University | Turker M.,Pakmaya | Dincer H.,Kocaeli University
Drying Technology | Year: 2010

A control system was designed using adaptive neuro-fuzzy inference system (ANFIS) for industrial-scale batch drying of baker's yeast. The temperature and flow rate of inlet air were considered as the manipulated variables to control the temperature and dry matter of the product, respectively, resulting in two adaptive fuzzy controllers. The membership functions for all inputs were adjusted by a hybrid learning algorithm. The database used in this work comprises large quantities of industrial-scale data (about 570 batches) obtained under different working conditions over one year. This database was used for learning and testing phases of the ANFIS controller. The performance of the proposed controller demonstrates the effectiveness and potential of the proposed ANFIS-based controller. © 2010 Taylor & Francis Group, LLC. Source


Hocalar A.,Pakmaya | Turker M.,Pakmaya
Biochemical Engineering Journal | Year: 2010

The specific growth rate should ideally be maintained at maximum oxidative growth rate in order to maximize biomass yield and productivity in fed-batch yeast fermentations. However, a more conservative approach is adopted in industry where specific growth rate is kept below the critical value to prevent the accumulation of overflow metabolite ethanol by using predetermined feeding profiles. In this work, biomass growth is maintained just above maximum oxidative growth rate by regulating ethanol concentration in the fermenter. The state feedback linearizing control strategy is developed and applied to the technical scale fed-batch yeast fermentations. The proposed control algorithm is constructed on the reliable state estimation algorithm developed previously and the biomass concentrations and ethanol measurements are then used in the control algorithm. The ethanol concentration is successfully controlled at fixed and time varying set values. By this approach the specific growth rate is controlled just over the critical value by regulating minimal ethanol concentration in order to maximize the biomass productivity. © 2010 Elsevier B.V. Source


Dogan E.C.,Kocaeli University | Turker M.,Pakmaya | Dagacsan L.,Pakmaya | Arslan A.,Kocaeli University
Biotechnology and Bioprocess Engineering | Year: 2012

Bio-oxidation of sulfide under denitrifying conditions is a key process in the treatment of gas and liquids that are contaminated with sulfide and nitrite. A lab-scale continuous flow stirred tank reactor (CFSTR) was operated with nitrite as the electron acceptor for the evaluation of the effects of loading rates, hydraulic retention time (HRT) and substrate concentrations on the performance of the autotrophic denitrification process. The influent sulfide concentration was maintained at 0.16 kg/m3 and the HRT was decreased from 8.4 to 2 h and for the entire study period, the sulfide removal efficiency was above 80% for the loading rates that ranged from 0.47 to 2.16 kg S -2/m3day. However, lower influent loading of NO 2 --N that correspond to the stoichiometric ratios was used and the nitrite removal efficiency was close to 100%. The stoichiometry of sulfide oxidation was calculated by assuming the different end-products based on thermo-dynamic approach. We compared the calculated values with experimental yield values. © 2012 The Korean Society for Biotechnology and Bioengineering and Springer. Source


Hocalar A.,Pakmaya | Turker M.,Pakmaya | Karakuzu C.,Bilecik University | Yuzgec U.,Bilecik University
ISA Transactions | Year: 2011

In this study, previously developed five different state estimation methods are examined and compared for estimation of biomass concentrations at a production scale fed-batch bioprocess. These methods are i. estimation based on kinetic model of overflow metabolism; ii. estimation based on metabolic black-box model; iii. estimation based on observer; iv. estimation based on artificial neural network; v. estimation based on differential evaluation. Biomass concentrations are estimated from available measurements and compared with experimental data obtained from large scale fermentations. The advantages and disadvantages of the presented techniques are discussed with regard to accuracy, reproducibility, number of primary measurements required and adaptation to different working conditions. Among the various techniques, the metabolic black-box method seems to have advantages although the number of measurements required is more than that for the other methods. However, the required extra measurements are based on commonly employed instruments in an industrial environment. This method is used for developing a model based control of fed-batch yeast fermentations. © 2011 ISA. Published by Elsevier Ltd. All rights reserved. Source


Turker M.,Pakmaya | Baspinar A.B.,Yildiz Holding | Hocalar A.,Pakmaya
Journal of Chemical Technology and Biotechnology | Year: 2012

Background: Hydrogen sulphide (H 2S) present in biogas can be oxidized to elemental sulphur (S 0) or sulphate (SO 4 2-) using nitrate and nitrite. Both nitrate and nitrite are normally available in most wastewater treatment plants and could be used to oxidize H 2S depending on the molar loading ratio of wastewater and biogas. A control approach is required in order to minimize the fluctuations in inlet and outlet H 2S concentrations in biogas, and the oxidation potential of the wastewater used. Results: A control scheme has been developed for biogas desulphurization using oxidation reduction potential under industrial conditions. The redox potential was maintained at about + 50 to + 100 mV in the activated sludge plant to monitor the performance of the nitrification process. The redox potential in the bioscrubber was related to sulphide removal from biogas. More than 90% of the hydrogen sulphide was removed from the biogas. Conclusion: The oxidation reduction potential can be used as a key parameter for monitoring and controlling biogas cleaning. Fluctuations of the inlet H 2S concentration in biogas can be compensated by manipulating the flowrates of wastewater used in order to achieve consistent and desired H 2S concentrations in treated biogas. © 2011 Society of Chemical Industry. Source

Discover hidden collaborations