MODELING DISSOLVED OXYGEN (DO) AND BIOLOGICAL OXYGEN DEMAND (BOD) ON LOW FLOW CONDITIONS USING ARTIFICIAL NEURAL NETWORK (ANN) IN CIKAPUNDUNG RIVER

• Main Author: NASRULLAH, IRFAN
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/28122
• Institution: Institut Teknologi Bandung

Low flow condition is an extreme condition in the river that occurs when water capacity comes into minimum so the river pollutant receive load is increased so that the river water quality worsened. In hydrology, low flow statistics 7Q10 hydrological method used to determine low flow for limiting waste disposal permitted. In the river modeling research and management, indicator of the river water quality that used are Dissolved Oxygen (DO) and Biochemical Oxygen Demand (BOD) both have an important role for measuring instrument that affect in river management strategy. In this research, Cikapundung River is modeled using computational numerical modeling with Artificial Neural Network (ANN) method so that it can be known the low flow value relating with the quality of DO and BOD in river flow. The scenario simulation using feed forwardback propagation (FFBP) algorithm with two layers configuration and hidden neuron with tangent-sigmoid (tansig) in first layer and pure-linier (purelin) transfer function in second layer. The programming of model runs on Matlab 2015b.