A Multi-Objective Best Compromise Decision Model for Real-Time Flood Mitigation Operations of Multi-Reservoir System

• Published in: Water resources management Vol. 30; no. 10; pp. 3363 - 3387
• Main Authors: Jia, Benyou, Simonovic, Slobodan P., Zhong, Pingan, Yu, Zhongbo
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.08.2016; Springer Nature B.V
• Subjects: Algorithms; Atmospheric Sciences; case studies; China; Civil Engineering; Decision making; Earth and Environmental Science; Earth Sciences; Environment; Flood control; Flood damage; Floods; Flow control; Freshwater; Fuzzy sets; Geotechnical Engineering & Applied Earth Sciences; Heuristic; Historic floods; Hydraulics; Hydroelectric power; Hydrogeology; Hydrology; Hydrology/Water Resources; Linear programming; Mathematical models; Optimization algorithms; Optimization techniques; Real time; Reservoir management; Reservoirs; River basins; Studies; utility functions; watersheds; China; China, People's Rep; Differential evolution; Multi-objective optimization; Multi-reservoir system; Real-time reservoir system operations; Fuzzy analytic hierarchy process; Flood mitigation
• ISSN: 0920-4741; 1573-1650
• DOI: 10.1007/s11269-016-1356-0

To address the decision-making problem for real-time multi-objective flood operations in multi-reservoir system, this paper develops a multi-objective best compromise decision model (MoBCDM). Utility function is used to quantitatively express the preference of decision maker, and also fuzzy analytic hierarchy process (FAHP) and segmentation and averaging (Seg/Ave) are adopted together with the preferences of decision participants (hydrologist and reservoir manager) to convert the problem into a scalar optimization. The differential evolution (DE) algorithm is implemented for obtaining the best compromise solution. The multi-objective flood operation problem in Shiguan River Basin (in China), which contains two reservoirs and three flood control points, is used as a case study. The analyses are performed to compare four historical flood operations scenarios, this model and current operating rules. The results of the analyses show that the MoBCDM outperforms all operational scenarios in terms of peak flow reduction at three flood downstream control points. In addition, the MoBCDM execution is very efficient in real-time implementation, and also weighting coefficients for the use by the MoBCDM can get high resolution calculated by FAHP.