Determination of water quality parameters and nutrient level with an Adaptive Neuro- Fuzzy Inference System

Abstract

In this research, the physico-chemical water quality parameters and the effect of climate changes on<br />water quality is evaluated. During the observation period (5 months) physico-chemical parameters<br />such as water temperature, turbidity, saturated oxygen, dissolved oxygen, pH, chlorophyll-a, salinity,<br />conductivity, and concentration of total nitrogen (nutrient level) as main pollutant factor have been<br />measured in Iran from September to February 2013 in the Amirkabir dam area. Moreover, an<br />adaptive neuro fuzzy inference mechanism (ANFIS) is designed for the sake of modeling and<br />prediction. In order to learn the proposed ANFIS mechanism a Quantum behave particle swarm<br />optimization (QPSO) is employed. The proposed ANFIS architecture has nine-input and one output<br />in which the physico -chemical parameters of water and total nitrogen have been considered as input<br />and output of the proposed ANFIS, respectively. In this paper to reduce the noise and measurement<br />errors a wavelet transform strategy is utilized.