A successful candidate strategy with Runge-Kutta optimization for multi-hydropower reservoir optimization. (15th December 2022)

• Record Type: Journal Article
• Authors: Chen, Huiling; Ahmadianfar, Iman; Liang, Guoxi; Bakhsizadeh, Hedieh; Azad, Babak; Chu, Xuefeng
• Is Part Of: Expert systems with applications. Volume 209(2022)
• Publication Date: 2022-12-15
• Subjects: Runge-Kutta -- Optimization -- Hydropower -- Multi-reservoir -- Successful candidate strategy -- RUN -- Runge Kutta Optimization -- Metaheuristic; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2022.118383
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: Successful candidate strategy coupled with Runge-Kutta optimization (ScsRUN) is proposed. Successful candidate strategy is added to boost global and local search mechanism. A novel enhanced solution quality (ESQ) is adapted to break free from local positions. ScsRUN is evaluated over 29 benchmark functions and a complex real world problem. ScsRUN's superiority over advanced algorithms has been shown. Abstract: The hydropower problems are non-linear and non-convex in nature; therefore, optimizing the operations of multi-hydropower plants in a multi-reservoir system is always challenging and complicated. In order to develop an appropriate optimization technique to address such difficulties, an efficient search strategy with a high capacity to move from exploration to exploitation in the feasible domain is required. This paper provides a successful candidate strategy combined with Runge-Kutta optimization (ScsRUN) to quickly, accurately, and reliably optimize multi-reservoir hydropower problems. Specifically, the successful candidate strategy is used to improve the stability between exploration and exploitation phases; a modified version of enhanced solution quality implemented in the original RUN (MESQ) is utilized to enhance the efficiency of solutions and break free from local positions; and sequential quadratic programming is performed as a robust local search method to accelerate convergence. The new optimization method developed in this study was evaluated by using 29 test functions and a real-world, complicated multi-reservoir problem. It was demonstrated that ScsRUN outperformed other advanced optimization methods in terms of efficiency and reliability. ScsRUN can be widely used to solve a variety of complicated problems. The last source codes of RUN algorithm is publicly available at https://imanahmadianfar.com .