Performance Comparison of Evolutionary and Nature Inspired Algorithms for Optimizing the Operation of Indira Sagar Reservoir

Abstract

Efficient reservoir operation is an important aspect of sustainable water resources management in monsoon-driven river basins such as India, where hydrological variability and competing water demands present complex operating challenges. The present study is aimed to evaluate the comparative performance of evolutionary and nature inspired metaheuristic optimization approaches for the optimal operation of Indira Sagar Reservoir. The reservoir operation problem is formulated as a constrained multi-objective optimization problem to minimize water-supply deficits and improve hydropower related performance under the constraints of storage, dead-storage, release, spillway-capacity and storage-change. The selected algorithms represent different population-based evolutionary and nature-inspired search mechanisms and are evaluated under a common modelling framework using representative hydrological years. Performance is assessed based on deficit reduction, convergence behavior, reliability, resilience, vulnerability, and sustainability indicators. The existing operating policy exhibited low time-based reliability (0.4627) and volumetric reliability (0.6854), while resilience reached 1.0000, indicating rapid recovery following detected failures but frequent shortage occurrence. The comparative results showed that CSA achieved the lowest mean deficit of 966.8 MCM among the evaluated approaches, compared with approximately 14,220.14 MCM for the existing policy, followed by IWO and CA. The observed ranking was CSA > IWO > CA > TV-EM-MOPSO > EM-MOPSO > MO-PSO. These findings identify CSA as the most promising candidate among the evaluated approaches for further development and validation of an improved operating policy for Indira Sagar Reservoir.