Planning & Management

Oral

396795 - Assessing multi-objective optimization algorithms for large-scale reservoir operating rules based on sensitivity analysis

Wednesday, June 6
8:30 AM - 10:00 AM
Location: Northstar B
Co-Authors: Pan Liu, Wuhan – Wuhan University; Xiaohui Lei, Beijing – China Institute of Water Resources and Hydropower Research; Zejun Li, Wuhan – Wuhan University; Hao Wang, Beijing – China Institute of Water Resources and Hydropower Research

With more reservoirs built in China, the application of large-scale reservoir system has gained more attention. But non-commensurable objectives and high dimensionality exist in the optimization of large-scale reservoir system. To solve these problems, three steps exist in the derivation of large-scale reservoir operating rules: (1) simplification of large-scale reservoir operating rules by aggregation-decomposition model, (2) identification of sensitive parameters by sensitivity analysis for dimensional reduction, and (3) three multi-objective optimization algorithms have been compared for deriving the large-scale reservoir operating rules. Non-dominated sorting genetic algorithm II (NSGAII) is the benchmark of three algorithms. Weighted crowding distance replaced the traditional crowding distance in weighted non-dominated sorting genetic algorithm II (WNSGAII). And the Weighted Multi-Objective Adaptive Surrogate Model Optimization (WMO-ASMO), embedded with WNSGAII, could reduce the computational cost and speed the searching process. The large-scale reservoir system of Xijiang River Basin in China are selected as a case study. Results indicate that: (1) WNSGAII performs better than the well-known NSGAII because of the weighted crowding distance, and (2) WMO-ASMO outperforms NSGAII and WNSGAII in effectiveness and efficiency with same original model runs. Therefore, WMO-ASMO is proved to be more efficient and could provide better Pareto frontier.

Jingwen Zhang

PhD student
Wuhan University

Presentation(s):

Send Email for Jingwen Zhang


Assets

396795 - Assessing multi-objective optimization algorithms for large-scale reservoir operating rules based on sensitivity analysis



Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Assessing multi-objective optimization algorithms for large-scale reservoir operating rules based on sensitivity analysis