APPLYING NEURAL NETWORKS FOR PREDICTING THE FACULTATIVE OLIGOTROPHIC BACTERIA IN TWO RESERVOIRS WITH DIFFERENT TROPHIC STATE

作者：Radojevic I^*; Comic L; Rankovic V; Ostojic A; Topuzovic M

来源：Journal of Environmental Protection and Ecology, 2013, 14(1): 55-63.

• facultative oligotrophic bacteria
• feedforward neural network
• modelling
• reservoir
• water quality

The objective of this study is to develop a feedforward neural network (FNN) models to predict the facultative oligotrophic bacteria in reservoirs of different trophic state. The data collected in the hydrobiological research of the Gruza eutrophic reservoir in which the heterotrophic bacteria is a dominant community and of the Grosnica mezotrophic reservoir where the facultative oligotrophic is a dominant bacteria group were used for creating models. The neural network models were developed using experimental data which is collected during 10 years. The input variables of the neural network for models are: water temperature, pH, electrical conductivity, total phosphate, nitrites, ammonia, chlorophyll a, dissolved oxygen, 5-day biochemical oxygen demand, chemical oxygen demand and number of heterotroph and are the same for the models. The Levenberg-Marquardt algorithm is used to train the FNN. The optimum FNN architecture was determined. Results of FNN models were compared with the measured data on the basis of mean absolute error (MAE) and mean square error (MSE). Comparing the modelled values by FNN with the experimental data indicates that neural network model provides accurate results.

• 出版日期2013