Modeling the influent and effluent parameters concentrations of the industrial wastewater treatment under zeolite filtration

The maximum pollution of wastewater discharged into a treatment plant located in the industrial town of Shiraz, Fars, Iran, has been used to define the optimum condition of zeolite filtration. This study aimed to focus on controlling water salinity used to water the green area that have measured the concentration fluctuations of the influent and effluent parameters, e.g., Chemical Oxygen Demand (COD), Electrical Conductivity (EC), Temperature, Total Suspended Solids (TSS), Total Dissolved Solids (TDS), and pH. In the first step, we tried to obtain a model using general regressions, e.g., the linear or nonlinear approaches. Here, it was assumed that the reason for failing these approaches on modeling was an unsteady operation of the zeolite behavior on wastewater purgation, which could be trained by a lag-time associated with time series analysis, e.g., ARIMA. The second step was to use the Fuzzy Inference System, FIS, for finding the relation between the COD concentration value of the effluent with the concentrations of TDS, TSS, and EC of influent. Finally, to achieve a model for simulating the removal of COD against other parameters, the Adaptive Neural Fuzzy Inference System (ANFIS) was used. The vague and intangible zeolite behavior led to form the fuzzy inference systems, and there thus resulted in good distribution between estimated and observed data. Because the regression coefficient on comparing the simulated distribution of COD concentration of daily effluent with the line of 1:1 was 0.764. The results revealed that two extreme points existed for the fluctuations of COD concentration effluent led to its lowest or highest levels due to the concentration fluctuations of Na and NH4 + , where its removal efficiency has maximum or minimum without the steady-state conditions. This study observed that the role of zeolite was the disruption of ion-exchange stability by constantly arousing their interaction and has the reason of failure iteration to improve the regression coefficient of linear and nonlinear functions via the curve fitting approach and clustering analysis with and without normalized data. The fuzzy set theory was very helpful in this condition, such showed the soluble solids concentration (e.g., 1877 mg/l) on the discharge wastewater in the industrial treatment plant has a threshold value whose an important role in the zeolite filtration effectiveness.