Assessment of groundwater corrosivity in Hamedan Province, Iran using an adaptive neuro-fuzzy inference system (ANFIS)

作者：Abyaneh Hamid Zare^*; Varkeshi Maryam Bayat; Mohammadi Kourosh; Howard Ken; Marofi Safar

来源：Geosciences Journal, 2011, 15(4): 433-439.

DOI：10.1007/s12303-011-0032-x

• corrosion
• groundwater
• Langelier index
• Puckorius index
• ANFIS algorithm

The Langelier Saturation Index (LSI) and Puckorius Scaling Index (PSI) were used to study the incrustation/corrosion potential of groundwaters in Hamedan Province, Iran. The LSI and PSI indices correlated strongly with total dissolved solids (TDS) (mgL(-1)), pH and HCO (3) (-) (mmolL(-1)) as revealed by determination coefficients of 0.90 and 0.99 for LSI and PSI, respectively. Subsequently, a trained adaptive neuro-fuzzy inference system (ANFIS) was deployed to predict the corrosion behavior of water in unsampled or partially sampled locations using available total dissolved solids (TDS), pH and HCO (3) (-) as input data. Excellent agreement between the ANFIS simulations and the indices determined using conventional LSI and PSI techniques confirms that modeling using the ANFIS approach will allow water corrosion potential to be determined reliably and inexpensively using only TDS, pH, and HCO (3) (-) data.

• 出版日期2011-12