BOD-COD-T0CAI-Based for Prediction Model for Carbon Removal in Full-Bed Configuration of Biological Aerated Filters Utilizing Neural Network

Existing Total Organic Carbon (TOC) in wastewater cause to decrease amount of oxygen and so, It causes to death the sea organisms.It can be analyzed by amount of chemical oxygen demand (COD) and biochemical oxygen demand (BOD). It is because, BOD andCOD experimental outputs have used for evaluation of microorganism in a partial bed reactors to find the percent of existing carbonin wastewater thatit depends on Influenced Totally Organic Carbon (ITOC), Effluent Totally Organic Carbon (ETOC) and organicloading rate (OLR). After the start-up, the reactor has parallel operated at the same hydraulic and organic loading rates. For each steploading in this research, we analyzed the amount of the influent and effluent of thereactor during of unsteady andSteady-stateoperations period by measurement TOC in influent and effluent. The removal rate for each loading was calculated according to themean TOC removal efficiencies and the mean OLRs applied. As we know, experimental work is difficult and so, using a nonexperimentalmethod to collect output data with minimum error and maximum correlation help us to be faster. Artificial neuralnetwork (ANN) is a branch of intelligent network that is applied in current study to predict the carbon removal percentage in partialbed reactor. ITOC, ETOC and OLR are input and carbon removal is output. 300 data used for training and 30 date used for testing.The best network selected base on Root Means Squares Error (RMSE) equal 0.07% and high correlation coefficient equal 0.973.