A semi-infinite analysis-based inexact two-stage stochastic fuzzy linear programming approach for water resources management

A semi-infinite analysis-based inexact two-stage stochastic fuzzy linear programming approach is proposed for water resources management problems. Semi-infinite analysis is incorporated into the modelling framework to handle infinite α -cuts to the membership functions of fuzzy parameters. Compared to the existing methods in which only limited α -cuts are divided under an assumption that these could represent all fuzzy information of the function, the proposed approach could be more reliable in terms of full consideration of fuzzy membership functions. The obtained optimal results from this approach present robustness according to a case study. By assigning different confidence levels, five groups of optimal solutions are obtained. It is indicated that higher confidence levels would relate to more reliable allocation strategies, as well as higher and more stable system benefits. However, with the increases in violation risks, the intervals of the optimal system benefit become wider, reflecting the impact of high risks combined with high benefits.