Optimal multipurpose reservoir operation planning using Genetic Algorithm and Non Linear Programming (GA-NLP) hybrid approach

Abstract

The major issue in water resource planning and management is determination of the optimal operating policy of a reservoir. In this study, GA-NLP hybrid approach has been applied. Genetic Algorithm (GA) is able to find the near optimal solution easily but for finding global optimal solution it will take too many function evaluations, so hybrid approach is adopted. GA is based on the principle of natural selection, derived from theory of evolution and it is popular for solving optimization problems. Non Linear Programming (NLP) is based on second derivatives of the Lagrangian function. To show practical utility, this approach is applied to an existing reservoir namely, Nagarjuna sagar Reservoir system in India. Maximizing sum of Irrigation releases into left and right main canal from the reservoir while maximizing annual power production is considered as the objective function. The constraints considered for this model are bounds for releases, storage capacity, canal capacity, and power generation capacity. The proposed model derives the optimal rule curves and release policies of the reservoir for different inflow scenario’s and for different priorities. Hence, based on the present case study it can be concluded that GA-NLP hybrid approach has the capability to perform efficiently, if applied in real world reservoir operation problems.