Making Emergency Decision Based on Fuzzy and Uncertain Multiple Attribute Decision


Student thesis: Doctoral Thesis

View graph of relations



Awarding Institution
  • Kim Meow LIEW (Supervisor)
  • Jinhua Sun (External person) (External Supervisor)
Award date20 Jul 2021


Recently, various types of emergencies have occurred frequently around the world, which have not only brought about tremendous casualties and property losses, but also posed great negative impacts on economic development, social stability and public safety. On the other hand, with the development of economy and society, although emergency management (EM) and emergency alternative systems have been improved, and emergency response capacity (ERC) has been strengthened, there are still many problems and lack of scientific evaluation, decision and optimization methods. Consequently, emergency decision making (EDM) has attracted more and more attentions from national governments and even the entire international community. For the whole cycle of EM, no matter EDM is carried out before emergencies, during emergencies or after emergencies, it inevitably always requires reasonable and effective solutions, which is of great significance to reduce potential losses and adverse effects caused by emergencies. However, in fact, because of the lack of information, uncertainty and complexity of emergencies, it is difficult to address EDM problems using conventional management methods. Fuzzy multiple attribute decision making (MADM) methods are regarded as one effective way for the determination of the optimal solutions in EDM. Hence, many studies have been devoted to EDM methods based on fuzzy MADM techniques, and obtained abundant outcomes. Nevertheless, there are still some shortcomings in EDM methods within the framework of fuzzy MADM, which fail to cope with the practical EDM problems fully and effectively. In response, this thesis aims to investigate new EDM methods for the whole cycle of EM based on the fuzzy and uncertain MADM.

First, for the EM optimization problem before emergencies, an EM optimization method based on interval 2-tuple linguistic group consensus Decision Making Trial and Evaluation Laboratory (DEMATEL) technology is developed. The diversity, ambiguity and uncertainty assessments about the direct relationships among influential factors of EM are flexibly and exactly modeled by interval 2-tuple linguistic information. A consensus reaching algorithm is incorporated into the suggested approach, which enables an acceptable degree of consensus among experts concerning the interval linguistic evaluations about direct influence relations among factors. The probability of belonging to cause group or effect group for each factor is calculated, and then the construction of a probability-based cause-effect relationship diagram is proposed which can offer more beneficial information about the complex causal structure of EM system. Furthermore, an optimization process exploration algorithm is developed, by which both the best optimization sequence and optimization combination of factors can be derived for some situations restricted by available resources. An illustrative example is provided, and discussion analyses including sensitivity, validation and superiority analyses are conducted, which demonstrate the effectiveness and superiority of the proposed method.

Then, for the emergency alternatives collaborative selection problem during emergencies, a multiple department emergency alternatives collaborative selection method based on interval 2-tuple linguistic BWM-TODIM is developed. The evaluations provided by decision makers (DMs) are represented by interval 2-tuple linguistic information to flexibly reflect and exactly handle the underlying diversity, vagueness and uncertainty. Based on DMs’ evaluations, the individual performance evaluations about individual criteria and collaborative performance evaluations about collaborative criteria of multi-alternative combinations composed by different department alternatives are determined. The Best Worst Method (BWM) is extended into interval 2-tuple linguistic environment and the group decision making is taken into account in an interval fuzzy mathematical programming model, which can furnish a simple and reliable way for determining the weights of criteria. Besides, an interval 2-tuple linguistic TODIM (an acronym in Portuguese of interactive and multi-criteria decision making) method is proposed by considering DMs’ psychological behaviors and both the gain degree and loss degree of one alternative relative to another are simultaneously computed, which enable more practical and accurate decision results. The multi-stage dynamic decision process of the proposed method is also suggested that can suit the development of emergency and contribute to timely and effective response measures. An illustrative example together with discussion analyses containing sensitivity, validation and superiority analyses are carried out, through which the effectiveness and advantages of the proposed approach are demonstrated.

Finally, for the ERC evaluation problem after emergencies, a multi criteria comprehensive evaluation approach for ERC with interval 2-tuple linguistic Analytic Hierarchy Process (AHP) and interval aggregation operators is developed. Experts’ diversified, fuzzy and uncertain assessment information is flexibly and accurately processed by virtue of interval 2-tuple linguistic model. The AHP method is extended into interval 2-tuple linguistic information environment, where the interval 2-tuple linguistic preference relation (ITLPR) and its multiplicative consistency are defined. An iterative algorithm to improve the consistency level of an ITLPR is proposed which enables automatic consistency improvement. Based on the improved ITLPRs, criteria weights in the form of interval 2-tuple linguistic variables are derived by the normalizing rank summation method. In individual decision information collecting processes, both subjective and objective weights of experts are considered that present a more reasonable and practical strategy to determine experts’ weights. Moreover, some interval 2-tuple interval weighted aggregation operators are developed which can suit the associated weights in interval 2-tuple linguistic form and yield comprehensive final evaluation results. By taking advantage of the amount of positive information and reliability of information, a new comparison method for interval 2-tuples is proposed and can synthetically compare the overall performances of ERC and the weighted performances on sub-criteria of evaluation objects, and further give clear suggestions for improving ERC. A numerical illustration and discussion analyses including sensitivity, validation and superiority analyses are conducted, by which the feasibility and superiority of the proposed method are elucidated.

    Research areas

  • Emergency decision making, Multiple attribute decision making, Interval 2-tuple linguistic, Emergency, Emergency management, Emergency alternative, Emergency response capacity