Gradient Evolution Optimization Algorithm to Optimize Reservoir Operation Systems

doi:10.1007/s11269-018-2122-2

Return

Gradient Evolution Optimization Algorithm to Optimize Reservoir Operation Systems

2018-10-26

PRE

Arvin Samadi-Koucheksaraee

Iman Ahmadianfar *

Omid Bozorg‐Haddad

Seyed Amin Asghari Pari

DOI：10.1007/s11269-018-2122-2

Original

Original request for help

Save

Abstract

En 中文

Population growth, environmental destruction, and climate change have all led to water scarcity on the available water resources. In this regard, reservoir systems have an important role to manage water resources. Thus, it is essential to optimize the management of water resources. Optimizing reservoir systems involves complications such as nonlinear functions, large number of sizing variables and numerous constraints. To solve complicated optimization problems, meta-heuristic optimization algorithms are reliable and powerful methods. Hence, the present paper applies gradient evolution (GE) algorithm to optimize reservoir operation systems. This algorithm is extracted from a gradient-based optimizer. In fact, the main novelty of this study is the application of GE algorithm to optimize single- and multi-reservoir systems. Accordingly, the GE is employed to optimize a four-reservoir system, the Khersan-1 reservoir and the Dez reservoir in Iran. The results confirm the high capacity of the GE to optimize the single and multi-reservoir systems as it can obtain solutions 99.99, 96 and 94% of global optimum for the four-reservoir, Khersan-1 reservoir and Dez reservoir operation problems respectively. The results of the GE are compared with those solutions calculated with linear programming (LP), non-linear programming (NLP) and genetic algorithm (GA), which corroborate the superior ability of GE to reach global optimum solution of reservoir operation systems.

Keywords:

Optimization

Reservoir operation

Gradient evolution

Hydropower generation

Irrigation supply

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.