Predict Acute Phytotoxicity in Petroleum Contaminated Soil using Artificial Neural Network and Whale Optimization Algorithm

Abstract

Pollutants in the environment, particularly those derived from total petroleum hydrocarbons (TPH), affect soils chemically, biologically, and physically in an extremely complicated manner. Here, we examine this impact by modelling the acute phytotoxicity effects of TPH on soils. To achieve this goal, we have designed a predicted model using artificial neural network (ANN). The ANN algorithm performance depends on the hyper parameters used in it. Thus, determining the optimal hyper parameter values helps to enhance the prediction model. To achieve this goal, in this paper, we have done the hybridization of ANN and whale optimization (WO) algorithm. The whale optimization algorithm is a bio-inspired algorithm and successfully applied in different applications to determine optimal global solution. Therefore, in the proposed method, whale optimization algorithm is deployed for determine hyper parameter values of ANN. Further, MATLAB software is used for simulation purposes. The prediction model validation is done various parameters such as root mean square error (RMSE), mean absolute error (MAE), and mean square error (MSE). The result shows that the proposed method achieves lowest value of these parameters over the existing algorithms. This reflects that the proposed method is superior for predict acute phytotoxicity in petroleum contaminated soil and can be deployed for real-time applications.