Artificial neural network and multi-criteria decision-making methods for the remediation of soil oil pollution in the southwest of Iran

Oil pollution has notable effects on soil properties. There are different methods available to remove and decrease oil spill in soil such as the single or the combined use of physical, chemical, and biological methods. The main challenge in the remediation of oil-polluted soils is to find the most efficient methods to remove the contamination. The use of artificial intelligence such as artificial neural network (ANN) and multi-criteria decision-making (MCDM) methods, such as technique for order of preference by similarity to ideal solution (TOPSIS) and fuzzy methods, may be a useful way to investigate the efficiency of remediation methods. Accordingly, the objective of the present study was to find the most efficient method for the remediation of soil oil spill using ANN, TOPSIS, and fuzzy method. The study was carried out in an oil pumping facility in the southwest of Iran in 2017. Soil samples were collected from seven regions around the study area and their physicochemical properties were determined. The main criteria for ranking different methods were on the basis of intuitive methods. A radial basis function (RBF) type of ANN was used. The biological extraction method was the most efficient one. There was a perfect consistency between RBF outputs and the prioritized results by TOPSIS. In addition, a comparison of the results with fuzzy method revealed that ANN (r = 0.931 and MSE = 0.82) leads faster to a decision-making strategy for the remediation of polluted soils. However, using the network for making a decision requires a complete dataset.

Automated summary

Oil pollution has significant effects on soil properties. Various methods, including physical, chemical, and biological methods, can be used to remove and reduce oil spill in soil. The study aimed to find the most efficient method for soil oil spill remediation using artificial neural network (ANN), technique for order of preference by similarity to ideal solution (TOPSIS), and fuzzy method. The results showed that the biological extraction method was the most efficient, and there was consistency between the ANN and TOPSIS results. Comparatively, using ANN led to a faster decision-making strategy for the remediation of polluted soils.