水资源与环境约束的PROMO-HRA优化研究在WRM刊出

Uncertainty-based Multi-objective Decision Making with Hierarchical Reliability Analysis under Water Resources and Environmental Constraints
Water Resource Management

Feifei Dong  • Yong Liu  • Han Su • Zhongyao Liang  • Rui Zou • Huaicheng Guo 

Abstract Rapid urbanization and population growth have resulted in worldwide serious water shortage and environmental deterioration. It is then essential for efficient and feasible allocation of scarce water and environment resources to the competing users. Due to inherent uncertainties, decision making for resources allocation is vulnerable to failure. The scheme feasibility can be evaluated by reliability, representing the failure probability. A progressive reliability-oriented multi-objective (PROMO) optimal decision-making procedure is proposed in this study to deal with problems with numerous reliability objectives. Dimensionality of the objectives is reduced by a top-down hierarchical reliability analysis (HRA) process combining optimization with evaluation. Pareto solutions of the reformulated model, representing alternative schemes non-dominated with each other, are generated by a metalmodel-based optimization algorithm. Evaluation and identification of Pareto solutions are conducted by multi-criteria decision analysis (MCDA). The PROMO procedure is demonstrated for a case study on industrial structure transformation under strict constraints of water resources and total environmental emissions amounts in Guangzhou City, South China. The Pareto front reveals tradeoffs between economic returns of the industries and system reliability. For different reliability preference scenarios, the Pareto solutions are ranked and the top-rated one was recommended for implementation. The model results indicate that the PROMO procedure is effective for model solving and scheme selection of uncertainty-based multi-objective decision making.
Keywords Water resources allocation• Stochastic programming • Tradeoff analysis • Multi-criteria decision analysis • Multi-objective evolutionary algorithm • Optimization