Optimizing reservoir operation to avoid downstream physical habitat loss using coupled ANFIS- metaheuristic model

The present study proposes a novel optimization framework of the reservoir operation in terms of optimizing storage benefits while alleviating downstream environmental impacts. We utilized adaptive neuro fuzzy inference system as the habitat model to assess physical habitat loss at downstream of the reservoir. Furthermore, we applied metaheuristic algorithms including genetic algorithm, particle swarm optimization, invasive weed optimization, differential evolution algorithm, shuffled frog leaping algorithm, gravity search algorithm, atom search algorithm and bat algorithm to optimize reservoir operation. The developed objective function minimizes difference between physical habitat suitability in the natural flow and the optimal release in the simulated period. Based on the results in the case study, The Nash–Sutcliffe model efficiency coefficient demonstrated that the proposed ANFIS model is reliable to assess physical habitats. Hence, it could be used in a coupled simulation-optimization framework to optimize environmental flow regime downstream of a reservoir. Moreover, decision-making system indicated that gravity search algorithm is the best solution for optimization in terms of reliability, vulnerability and mean absolute error. The proposed method can be a robust framework to minimize physical habitat loss at downstream of the reservoirs.