Large scale reservoir operation through integrated meta-heuristic approach

Reservoir optimization models are often large-scale, having a complex, nonlinear, multi-dimensional structure, which poses a challenge for classical methods for their solution. This has encouraged the researchers to focus on Meta-heuristic which due to their flexible and adaptive nature have been successful in solving a plethora of real-life optimization problems. This study brings forward an implementation and comparison of six well-known Meta-heuristics namely: Simulated Annealing, Genetic Algorithms, Particle Swarm Optimization, Differential Evolution, Artificial Bee Colony, and Cuckoo Search and an integrated version of these algorithms with dynamic programming for optimizing the reservoir operations policy. In addition, two adaptive variants of DE named: FCADE2 and SaDE are also considered for the comparison. The case study considered for Mula reservoir supplying water to Major Irrigation Project on River Mula (Godavari basin), Ahmednagar district, Maharashtra, India. The objective is to determine the optimum release policy for Mula reservoir. Performance of the algorithms is analysed on two data sets (1) single year and (2) 30-years.