Group Decision-making Approaches To Drought Disaster Risk Control Under Uncertain Environment

Since the 21 st century,drought disaster presents a further serious trend under the influences of global climate changes and intensified human activities,and it has become one of the important issues restricting the sustainable development of economy and society in various countries.China has been most affected by drought disaster,but her current drought relief and disaster reduction system is still seriously lagging behind.Under the major national strategic deployments,such as cooperative development of water resources-economy society-ecological environment,response to climate changes,our drought relief and disaster reduction work is facing new challenges and higher requirements.New concepts and ideas of drought disaster risk management have gradually been concerned and studied in China.Accordingly,this paper set sights on the existing problems and shortcomings in theory and practice of drought disaster risk control,followed the research frontiers in theory and method of risk management and group decision-making,and carried out research according to the overall thought of theoretical basis analysis,group decision-making methods construction and actual cases verification.This paper proposed a framework of group decision-making for drought disaster risk control with multiple participation,and desigend a stochastic EMD method for information extraction and aggregation in group decision-making.On these basis,this paper focused on the general problems in drought disaster risk control decision-making,and proposed these drought disaster risk control group decision-making methods based on uncertain linguistic information,uncertain mixed information and uncertain stochastic dynamic information.Finally,this paper selected the typical problems of agricultural drought disaster risk control in Henan province for empirical analysis,and formed the application model of these above methods.The research results could have theoretical significance for revealing the uncertainties of drought disaster risk control decision-making and constructing quantitative analysis models,play an important role in enriching and developing the theoretical and methodological system of group decision-making,and have practical application values for government to scientifically regulate and comprehensively deal with drought disaster risk.The main studies of this paper are listed as follows:(1)The multi-participant group decision-making framework for drought disaster risk control was proposed,and the uncertainty characteristics in group decision-making were analyzed.By analyzing the elements of drought disaster risk system and their interaction mechanism,this paper pointed out that the uncertainty of risk and the complexity of system were the decision-making environment for drought disaster risk control.Followed coordinated development concepts of water resources-economic society-ecological environment system,this paper constructed a multi-participant group decision-making framework for drought disaster risk control,and analyzed the feasibility and necessity of this research.And then,this paper analyzed the uncertainty characteristics in drought disaster risk control group decision-making from three aspects: the first one was the uncertainty such as randomness which reflected the inherent nature of drought disaster risk system;the second one was the uncertainty of group decision-making information caused by insufficient understanding,lack of information or knowledge;the third one was the uncertainty reflecting the group decision-making processes and methods.(2)Combining stochastic simulation and filtering analysis methods,the stochastic EMD(Empirical Mode Decomposition)approach to extracting and aggregating group decision-making information was proposed.Traditional group decision-making methods usually need expert weighting method to realize information aggregating of group decision-making,but the weighting methods had problems such as being subjectivity,lack of unified criteria and so on.This paper analyzed characteristics of group decision-making information such as diversity,non-fixed sequences and uncertainty,assumed that group decision-making information contained subjective and objective components and objective components were decisive in decision-making.And then,combining Monte Carlo stochastic simulation and EMD methods,this paper designed MC-EMD algorithm in order to adaptively extract and aggregate group decision-making information,thus to improve the objectivity and reliability of group decision-making results.The convergence of this algorithm was proved with law of large numbers,and the physical significance of this algorithm was also discussed.The feasibility and superiority of this method were compared and analyzed with numerical examples.(3)A group decision-making method for drought disaster risk control with uncertain linguistic group decision-making information was proposed by using the stochastic EMD method.This method mainly concerned the situation that group decision-making information were expressed with uncertain linguistic variable.This paper sorted out the existing linguistic evaluating scales,and designed multi-granularity linguistic information unified processing rules.To make full use of uncertain linguistic group decision-making information,combined basic concepts such as linguistic hybrid averaging(LHA)operator,numerical coverage and grey information representation,and membership function of fuzzy normal distribution,this paper respectively constructed three information aggregation methods with LHA operator,three-parameter interval grey number and normally distributed random variable.The comparative analysis of application example showed that the above methods were feasible and effective,and these methods can provide support for alternative demonstration of drought disaster risk control by multi-fields specialists.(4)A group decision-making method for drought disaster risk control under uncertain mixed group decision-making information was proposed by using the stochastic EMD method.This method mainly concerned the situation that group decision-making information were mixed with multivariate uncertainty expressions.This paper designed mixed decision-making information collecting method based on social investigation,summarized the commonly used representation methods of uncertain preference information,and proposed a transformation method of uniform normalized utility value.Furthermore,based on cloud model and the basic idea of Monte Carlo simulation,this paper proposed a normal cloud aggregation method under mixed uncertain group decision-making information,which can be used to remedy these deficiencies such as the information imperfect caused by random sampling of decision-makers,the information loss when processing information,and the conclusions inconsistent when making decision repeatedly.The reasonableness and effectiveness of this method was proved by an application example,and this method can provide support for alternatives decision-making of drought risk control for wide public participation.(5)A group decision-making method for uncertain stochastic dynamic drought disaster risk control was proposed by using the stochastic EMD method.This method mainly concerned the stochastic uncertainty of drought disaster risk evolution and the complexity of psychological and behavior of decision-makers,the reference point dependence principle and mixed uncertain group decision-making information were considered together.This paper proposed a general framework for group decision-making for drought disaster risk control.According to tri-reference point(TRP)theory,this paper set group ideal goal,group expected goal and group minimum requirement as reference points of risk-based group decision-making,then proposed the risk-based dynamic group decision-making information aggregation method based on cumulative prospect theory,Markov chain and orness measure theory.An application example was given to test and verify that the method was scientific and effective.This method can provide support for multi-stage dynamic group decision-making of drought disaster risk control.(6)Empirical analysis of agricultural drought disaster risk control in Henan province with those proposed group decision-making methods.This paper constructed a grey incidence analysis model with rate of change based on multi-factor cross-sectional data,and used it to identify key driving factors of agricultural drought vulnerability in Henan province.Aiming at the outstanding problems when preventing and controlling agricultural drought disaster risk in rain-fed agricultural areas and irrigated agricultural areas of Henan province,this paper constituted the evaluation indexes systems of regional applicability for policy-oriented agricultural insurance products and small-scale farmland water conservancy(SFWC)facilities management models,designed the evaluation questionnaire and social investigation outline to collect group decision-making information,and then,took Shanzhou district and Hua district as examples,used the proposed group decision-making methods to solve practical problems such as policy-oriented agricultural insurance products optimal delivery and SFWC facilities management models optimization,so as to form the application models of the proposed group decision-making methods.The above research results can provide support for general regional government not only to improve the efficiency of drought disaster risk control but also to enhance the democracy and scientific level of public decision-making.