Evaluation Of Effectiveness Of The Artificial Neural Network For Modeling The Cyanide Ions Adsorption From Aqueous Solution Using Zno@MOF-199 Nanoadsorbent

Cyanide is a by-product of various industrial chemical processes found in industrial effluents which must be treated before it is discharged into the environment. Industrial effluents containing cyanide is treated through different methods including physical, chemical and biological processes which are often too expensive.The aim of this study is to evaluate the application of Artificial Neural Network (ANN) in predicting the removal efficiency of cyanide ions from aqueous solutions by Zno@MOF-199 nano-adsorbent. The research data was collected based on laboratorial study of important parameters involved in the Levenberg-Marquardt Back-propagation Artificial Neural Network Model (BP-LM) for prediction of cyanide removal efficiency including pH range from 5 to 9, contact time of 30 to 90 minutes and temperature range from 25 to 45 ºC. Parameters such as pH, temperature, contact time, adsorbent weight and sample volume were considered as input and cyanide removal efficiency as output data. In the comparison of different models, statistical criteria of correlation coefficient and sum of squared errors (SSE) were applied. The obtained results, 0.985 for the correlation coefficient and 0.65 for the sum of squared errors, indicate a successful prediction of the network in modeling as well as the efficiency of neural network in predicting the removal efficiency of cyanide ions from the solution.