This study examines a general scheme of multicriteria decision-making under uncertainty based on quantitative and qualitative information. In its original form, the general scheme uses qualitative information at the final decision stage, only if the quantitative information is unable to generate unique solutions. Considering that many problems require objectives based on qualitative information at all decision stages, the results of Ramalho et al. (2019) can provide such solutions. However, the processing of qualitative information in Ramalho et al. (2019) is based on applying a very simple approach to aggregating individual preferences without analyzing any type of consensus information. Therefore, the present study sets out to show that representative combinations of initial data, states of nature or scenarios can be constructed by using qualitative information directly, based on aggregating individual preferences after achieving the necessary consensus. In particular, a new scheme for consensus construction is proposed. Using this scheme enables negative points inherent in traditional approaches to be avoided and uncertainty levels to be reduced when estimating coefficients of objective functions based on qualitative information. An illustrative example is presented to demonstrate the results that the new proposal can achieve.

本研究考察了基于定量和定性信息的不确定性下多准则决策的一般方案。仅当定量信息无法生成唯一的解决方案时，一般方案才能以其原始形式在最终决策阶段使用定性信息。考虑到许多问题在所有决策阶段都需要基于定性信息的目标，Ramalho等人的结果表明。（2019）可以提供这样的解决方案。但是，Ramalho等人对定性信息的处理。（2019）基于应用一种非常简单的方法来汇总个人偏好而无需分析任何类型的共识信息。因此，本研究着手显示初始数据的代表性组合，在达成必要的共识后，基于汇总个人偏好，可以直接使用定性信息来构建自然状态或情景状态。特别是，提出了一种共识构建的新方案。当基于定性信息估计目标函数的系数时，使用此方案可以避免传统方法固有的负点，并减少不确定性级别。给出了一个说明性示例，以演示新提案可以实现的结果。当基于定性信息估计目标函数的系数时，使用此方案可以避免传统方法固有的负点，并减少不确定性级别。给出了一个说明性示例，以演示新提案可以实现的结果。当基于定性信息估计目标函数的系数时，使用此方案可以避免传统方法固有的负点，并减少不确定性级别。给出了一个说明性示例，以演示新提案可以实现的结果。