Optimization model for integrated river basin management with the hybrid WOAPSO algorithm

Abstract

This work examines the effectiveness of a newly-developed optimization framework for river basin management. The proposed framework relies on the newly developed WOAPSO algorithm, which is a hybrid metaheuristic algorithm combining two conventional metaheuristic algorithms, namely the weed optimization algorithm (WOA) and the particle swarm optimization algorithm (PSO). Two case studies are presented in this study to evaluate the performance of the WOAPSO algorithm. The first case study consists of a ten-reservoir river basin example which compares the performance and reliability of the hybrid WOAPSO algorithm with that of linear programing (LP), non-linear programing (NLP), WOA, and the PSO algorithm. Results indicate the hybrid WOAPSO finds solutions meeting downstream water demands with 99.94% of reliability (with respect to the global optimum, as derived by LP) in the ten-reservoir system. It outperforms the WOA and PSO, which feature lower reliabilities than that achieved by WOAPSO. The second case study demonstrates failure of the conventional NLP optimization scheme in solving a real-world three-reservoir hydropower optimization problem which maximizes the efficiency index of hydropower production. The newly-introduced WOAPSO algorithm minimizes the objective function with superior efficiency compared with those of the WOA and PSO, in terms of the convergence rate and the achieved best values of the objective function. Furthermore, the WOAPSO is proven more reliable for solving complex multi-reservoir systems within the context of integrated river basin management than classic and evolutionary optimization algorithms.