Feilin Zhu

SMAA-based stochastic multi-criteria decision making for reservoir flood control operation

In reservoir flood control operation, candidate alternatives are generally evaluated, ranked and selected through multi-criteria decision making (MCDM) techniques, yet stochastic uncertainties both in the criteria performance values (PVs) and criteria weights (CWs) exist in the MCDM process. This paper extends the traditional MCDM methods to stochastic environments for reservoir flood control operation. The criteria PVs and CWs are treated as stochastic variables with certain probability distributions. The stochastic multicriteria acceptability analysis (SMAA) theory is introduced and the differences between conventional MCDM models and the SMAA-2 model are discussed. Methods for quantifying stochastic uncertainties in the criteria PVs are discussed and four kinds of CWs are proposed. Moreover, we define the concept of the risk of decision making errors and propose the corresponding quantitative calculation method. A three-stage MCDM procedure is recommended to guide decision makers to solve MCDM problems under stochastic environments. We apply the proposed methodology to a case study through Monte Carlo simulation to demonstrate its effectiveness and advantage. The results show that the proposed methodology can provide significant risk information for decision makers and improve the reliability of decisions for reservoir flood control operation.

Multi-criteria group decision making under uncertainty: Application in reservoir flood control operation

This paper proposes an innovative framework for solving stochastic multi-criteria decision making (MCDM) problems when uncertainties exist in criteria performance values (PVs) and criteria weights (CWs) simultaneously. Methods for quantifying uncertainties in criteria PVs and CWs are presented. We establish the SMAA-TOPSIS model by combining stochastic multicriteria acceptability analysis (SMAA) and technique for order preference by similarity to ideal solution (TOPSIS). The risk of decision making errors is proposed to assess the impact of uncertainties on MCDM. We develop the LHS-based Monte Carlo simulation algorithm and corresponding computer program for solving the SMAA-TOPSIS model. We also suggest a three-stage MCDM procedure for stochastic MCDM problems. We apply the proposed methodology to a flood control operation case study to demonstrate its applicability. Our results indicate that the proposed methods can provide valuable risk information and enable risk-informed decisions to be made with higher reliabilities. A novel SMAA-TOPSIS model for stochastic MCDM problems is proposed.The algorithm and program are designed for solving the SMAA-TOPSIS model.Methods for quantifying multiple uncertainties in MCDM are proposed.The risk of decision making errors is proposed to assess the effect of uncertainty.Helps decision makers make risk-informed decisions with higher reliabilities.

Bargaining Model of Synergistic Revenue Allocation for the Joint Operations of a Multi-Stakeholder Cascade Reservoir System

Given the institutional limitations of multi-stakeholders, exploring the synergistic revenue from the joint reservoir operations of a multi-stakeholder multi-reservoir system requires a synergistic revenue allocation mechanism to ensure a beneficial solution for multi stakeholders. This study established a synergistic revenue allocation model using bargaining game theory under the principles of equity, rationality, and efficiency. For the maximization the Nash product of gains in the utility of stakeholders and constraints on the feasibility of allocation plans considering all the possible formations of sub-coalitions, the analytic optimal solution of the bargaining model was derived using the first-order optimality condition. The optimal revenue allocation plan meets the conditions of the equal quasi-marginal utility function among stakeholders. The methodologies were applied to a hypothetical cascade reservoir system operated by multiple stakeholders. Compared with the revenue allocation plans obtained by a proportional rule method and the Shapley value method, the results corroborate that (1) the allocation plan of the bargaining model is jointly determined by the interval of the revenue range of each reservoir and the effectiveness of the sub-coalition constraints, indicating that the allocated synergistic revenue is positively correlated with the singleton contribution and team contribution on the total revenue of the grand coalition; (2) the difference in the plans obtained by the three methods is generally determined by the difference in equity definition; and (3) the synergistic revenue allocation plan obtained from the bargaining model is the highest homogenized among all reservoirs (stakeholders), which demonstrates that the revenue of those dominated stakeholders can be improved compared with other plans. The proposed methodologies provide new insights to guide benefit share decisions in multi-stakeholder reservoirs system.

Evaluation of global climate model on performances of precipitation simulation and prediction in the Huaihe River basin

Using climate models with high performance to predict the future climate changes can increase the reliability of results. In this paper, six kinds of global climate models that selected from the Coupled Model Intercomparison Project Phase 5 (CMIP5) under Representative Concentration Path (RCP) 4.5 scenarios were compared to the measured data during baseline period (1960–2000) and evaluate the simulation performance on precipitation. Since the results of single climate models are often biased and highly uncertain, we examine the back propagation (BP) neural network and arithmetic mean method in assembling the precipitation of multi models. The delta method was used to calibrate the result of single model and multimodel ensembles by arithmetic mean method (MME-AM) during the validation period (2001–2010) and the predicting period (2011–2100). We then use the single models and multimodel ensembles to predict the future precipitation process and spatial distribution. The result shows that BNU-ESM model has the highest simulation effect among all the single models. The multimodel assembled by BP neural network (MME-BP) has a good simulation performance on the annual average precipitation process and the deterministic coefficient during the validation period is 0.814. The simulation capability on spatial distribution of precipitation is: calibrated MME-AM > MME-BP > calibrated BNU-ESM. The future precipitation predicted by all models tends to increase as the time period increases. The order of average increase amplitude of each season is: winter > spring > summer > autumn. These findings can provide useful information for decision makers to make climate-related disaster mitigation plans.

Risk analysis for real-time flood control operation of a multi-reservoir system using a dynamic Bayesian network

Abstract This paper proposes a model for risk analysis of real-time flood control operation of a multi-reservoir system using a dynamic Bayesian network. The proposed model consists of three components: Monte Carlo simulations, dynamic Bayesian network establishing, and risk-informed inference for decision making. The Monte Carlo simulations provide basic data inputs for the dynamic Bayesian network establishing using the historical floods and operation models of the multi-reservoir system. The dynamic Bayesian network is built with expert knowledge and the relationships among the uncertainties. The component of risk-informed inference for decision making is to provide risk information about the operation schedules using the trained dynamic Bayesian network. We apply the proposed model to a multi-reservoir system in China. The results show that the proposed method has a capability for bi-directional inferences and can be served as a risk-informed decision-making tool under uncertainties in the real-time flood control operation of a multi-reservoir system.