Stochastic multi-criteria decision making based on stepwise weight information for real-time reservoir operation

Abstract

Reservoir operation is a typical combination of engineering and non-engineering measures, and plays an important role in sustainable water resources management. However, reservoir operators always work in uncertain and changing environments, and face various uncertainties arising from criteria weights (CWs) and performance values (PVs), which hinder the application of multi-criteria decision making (MCDM) methods in real-time reservoir operations. To address this issue, this paper proposes a stochastic MCDM framework based on stepwise weight information. The framework consists of five sub-modules/models: (1) formulate a stochastic MCDM problem and determine the stochastic decision matrix; (2) a weight aggregation method based on the game theory allied with feasible weight space is proposed to resolve conflict among multi-stakeholders and estimate the uncertainty of CWs; (3) a novel SMAA-FOM model, i.e., a combination of stochastic multi-criteria acceptability analysis (SMAA) and fuzzy optimization model (FOM), is developed to simulate uncertainties existed in PVs and CWs simultaneously and produces several descriptive measures for robust decision-making; (4) a risk assessment model is used to quantitatively evaluate the robustness and risk of decision-making; (5) a multi-stage decision-making procedure is proposed that enables a stochastic MCDM problem to be solved step by step within acceptable decision-making risks. The effectiveness and superiority of the proposed framework are verified via a multi-objective reservoir operation case study. The results indicate that the framework helps decision makers deal with stochastic MCDM problems with greater confidence and less psychological pressure, and enables a risk-informed decision can be made with higher reliabilities. The stochastic MCDM framework can be extended to other sustainability areas to solve decision-making issues with uncertainties.