Modeling and optimization of phycoremediation of heavy metals from simulated ash pond water through robust hybrid artificial intelligence approach

In thermal power plants, ash pond water is being polluted with heavy metals such as Lead (II), Nickel (II), Manganese (II), Cadmium (II), Chromium (VI), and so on, through leaching from coal ash and this can eventually contaminate the groundwater. In the present study, a consortium consisting of a cyanobacterium Synechococcus sp. and green algae Chlorella sp. was used for phycoremediation of such heavy metals from simulated ash pond water. The concentrations of metals, pH and days were varied during experimentation. An accurate data‐driven artificial neural network (ANN) model was developed with the experimental data to find a relation between the input variables of the phycoremediation process with the percentage removal of the pollutants and amount of biomass produced. The optimum ANN design and ANN algorithm were evaluated using an automatic exhaustive search technique. To maximize metal removal and biomass production, a hybrid artificial neural network‐simulated annealing (ANN‐SA) optimization technique was applied to determine optimal values of input parameters. The established modeling and optimization technique is generic and can be applied to any other experimental study. Several heavy metals such as Lead(II), Nickel(II), Manganese(II), Cadmium(II), Chromium(VI), and so on in the ash pond water contaminate the groundwater through leaching. In the present study, phycoremediation technique has been used to remove the heavy metals from simulated ash pond water. Artificial neural network (ANN) has been used to model the phycoremediation process, and hybrid artificial neural network‐simulated annealing (ANN‐SA) optimization technique has been used to optimize the process parameters and maximize the metal removal and biomass production.