Abstract: Various emergencies take place frequently worldwide, and therefore, emergency decision-making (EDM) has been the focus of governments and scholars. However, the complexity and uncertainty of the EDM problem make it difficult for decision-makers to make an effective decision. On the one hand, EDM is related to the vital interests of victims, thus the satisfaction of decision-makers and the public can affect the effectiveness of EDM. On the other hand, the evaluation of such satisfaction is often supported by heterogeneous data such as linguistic variables and fuzzy numbers. Therefore, an EDM method with utility and equity satisfaction under the fusion of multi-source heterogeneous data is proposed to solve the above problems. First, the EDM problem to be solved in this article is described (That is, in order to select the optimal alternative, how should we propose a scientific EDM method with utility and equity satisfaction under the fusion of multi-source heterogeneous data. In this method, different types of data such as linguistic variables and fuzzy numbers will be normalized to the form of interval numbers, and the functions of utility satisfaction and equality satisfaction will be further constructed based on new comparison method of interval numbers, based on which the comprehensive satisfaction and the ranking of alternatives will be obtained by using the weights of attributes which is obtained by an optimal model). Second, for the ease of understanding, the relevant notations are used to describe the decision information, and the decision data with various forms is represented by linguistic variables, intuitionistic fuzzy numbers and hesitant fuzzy numbers. Third, a method for fusing multi-source heterogeneous data is developed, and moreover, according to basic cases of interval number comparison, the method for comparing interval number and the comparison results in such cases is given. Then, by giving the subjective expectation interval of decision makers and the public on decision results in advance, the functions of utility satisfaction and equity satisfaction based on the prospect theory and the inequity aversion model are constructed respectively. Further, according to the principles of minimum deviation and maximum comprehensive satisfaction, an optimal model is established to obtain the weights of attributes, based on which the comprehensive satisfaction of alternatives is calculated. Finally, the COVID-19 is taken as an example to demonstrate potential application of the proposed method and the comparison with other approaches is carried out to demonstrate the effectives of the proposed method. The results shows that the proposed method can effectively solve the EDM problem, and compared with the exiting methods, the proposed method can not only better reflect the important role the utility and equity satisfaction play in selecting the optimal alternative, but also reflect the impact of decision-makers' preference for utility and equity satisfaction on decision results. Accordingly, the managerial implication of the study can be obtained as follows: In the actual decision-making process, it is usually difficult for the existing EDM methods to obtain the optimal decision result because the subjective psychological expectations of decision-makers and the public can affect their satisfactions to response result. Therefore, it is necessary for decision department to take utility satisfaction and equity satisfaction as important bases for decision-making, which can effectively avoid negative emotion caused by the dissatisfaction of the public to decision result as well as ensure the maximum of the decision utility satisfaction. In addition, in the decision process, the decision department should reasonably balance the relationship between the utility satisfaction and equity satisfaction of the alternatives according to the actual decision situation. For example, in small scale accidents such as urban public facilities accidents, the primary goal of decision-making is to reduce social losses caused by accidents as soon as possible, thus we should pay more attention to utility satisfaction of the alternatives in such a situation. In contrast, in large scale epidemic or earthquake disaster, because the decision process involves the problem of material distribution in different disaster areas and the beneficiaries are the victims in different disaster areas, the decision department should pay more attention to demands of the public on the basis of ensuring the maximum of the utility satisfaction, so as to prevent the negative emotion of the public caused by inequitable decision and its social influence. For future research, two directions will be further considered. One is that we have only considered the fusion problem of decision information which is represented as general forms such as linguistic variables and fuzzy numbers in this article, the future studies will consider the fusion problem of other types of decision data such as random variable. The other one is that the expectation intervals of decision-makers and the public are given in advance when constructing the functions of utility satisfaction and equity satisfaction, and future studies will consider the problem that how to construct the satisfaction function in the case of the uncertain expectation intervals.

Key words: multi-source heterogeneous data, emergency decision-making, utility satisfaction, equity satisfaction, prospect theory, inequity aversion model

摘要： 鉴于决策者和公众的满意度可影响决策效果,且其评估又受支撑于语言变量和模糊数等异构性数据,本文提出多源异构数据融合下考虑效用和公平满意度的应急决策方法。首先,描述该决策问题,并以语言变量和模糊数刻画形式多样的决策信息;其次,提出多源异构数据的融合方法和新的区间数比较方法;然后,分别基于前景理论和不公平厌恶模型构造方案的效用和公平满意度函数;进一步,构造优化模型获取最优属性权重,据此计算方案的综合满意度;最后,通过算例分析及与其它方法的对比说明所提方法的有效性。结果表明,与既有方法相比,所提方法不仅能够更好地体现效用和公平满意度在应急方案择优过程中的重要作用,还能够反映决策者对效用和公平满意度的偏好对决策结果的影响。

关键词: 多源异构数据, 应急决策, 效用满意度, 公平满意度, 前景理论, 不公平厌恶模型