| Current design of constructed wetland lacks essential parameters necessary to evaluate the removal of metals contained in water enters the system. Herein, for the first time, the artificial intelligence approach (Fuzzy Logic) is used to assess stochastic implication in the wetland systems. Bioavailable mercury was evaluated, using fuzzy logic approach, for different pH, initial concentration of inorganic mercury, and chloride concentration implied in the constructed wetlands. Fuzzy knowledge base was built based on results obtained from previous data: investigations that were performed in a greenhouse for floating plants, and previous computations for mercury speciation. Fuzzy Decision Support System (FDSS) used the knowledge bases to find out parameters that permit to generate the highest amount of bioavailable mercury for uptake by the floating plant. The fuzzy logic approach provided the required information on the capability of constructed wetland sediments to sorb mercury within the hazy conditions in the system. FDSS used the wetland knowledge bases to provide the final decision. Fuzzy knowledge bases were built manually on one stage and were generated genetically using Genetic Algorithm (GA) on the other stage, where the results in both stages show comprehensive and corresponding results of the soil performance in the system. The obtained information by the fuzzy logic approach supports into providing series of solutions for plant uptake and soil adsorption of mercury that represents the heavy metal removal from wastewater of the overall system, by which mathematical analyses and modeling were further established to set up constructed wetland design. The criterion of design assumes the removal process in wetlands is similar to a process that combines the treatment process in the attached growth reactor and the adsorption process in the granular activated carbon columns. (Abstract shortened by UMI.)... |
