Ultrasonic algae control system performance evaluation using an artificial neural network in the Doganci dam reservoir (Bursa, Turkey): a case study

ELMACI A., ÖZENGİN N., YONAR T.

DESALINATION AND WATER TREATMENT, cilt.87, ss.131-139, 2017 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 87
  • Basım Tarihi: 2017
  • Doi Numarası: 10.5004/dwt.2017.20810
  • Dergi Adı: DESALINATION AND WATER TREATMENT
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Sayfa Sayıları: ss.131-139
  • Anahtar Kelimeler: Artificial neural networks, Levenberg-Marquardt algorithm, Reservoirs, Ultrasonic algae control, CYANOBACTERIAL BLOOM CONTROL, FEEDFORWARD NETWORKS, WATER, PREDICTION, IRRADIATION, FLUCTUATIONS, ALGORITHM, RADIATION, DEPTH, LAKE
  • Bursa Uludağ Üniversitesi Adresli: Evet

Özet

Ultrasound is a well-established technology, but it has been applied only recently to control algal blooms. The main purpose of this study is to determine the appropriateness of field measurements for evaluating the performance of an ultrasonic algae control system using an artificial neural network (ANN) in the Doganci Dam Reservoir (Bursa, TURKEY). Within this study, data were obtained using the NeuroSolutions 5.06 model. Each sample was characterized using ten independent variables (time, total organic carbon (TOC), pH, water temperature (T-water), dissolved oxygen (DO), suspended solids (SS), the Secchi disc depth (SDD), open-water evaporation (E), heat flux density (H), air temperature (T-air), and one dependent variable (chlorophyll-a (Chl-a)). The correlation coefficients between the neural network estimates and field measurements were as high as 0.9747 for Chl-a. The results indicated that the adopted Levenberg-Marquardt back-propagation algorithm yields satisfactory estimates with acceptably low mean square error (MSE) values.