This paper studies the cost consensus model by considering the uncertain initial opinions and uncertain unit adjustment cost in group decision making. In the past consensus model based on optimization, the initial opinion and unit adjustment cost are usually assumed to be a crisp number for each expert. However, the speed of knowledge updating is often faster than people’s cognitive speed, it is difficult and impractical to ask experts to provide a clear initial opinion and determine the unit adjustment cost of each expert. In this paper, a new consensus approach is proposed to solve the above problems. First, a new distance measure is given based on interval-valued initial opinion, which retains the expert’s initial judgment and is consistent with most practical decision problems. Second, a linear analytical formula is given to reduce the computational cost of the piecewise function. Third, given the advantages of robust optimization in uncertain optimization, three robust cost consensus models are established to deal with the uncertain cost problem in consensus reaching progress. Finally, the proposed method is applied to P2P loan consensus, and sensitivity analysis and comparative analysis are presented.

本文通过考虑群体决策中不确定的初始意见和不确定的单位调整成本来研究成本共识模型。在过去基于优化的共识模型中，初始意见和单位调整成本通常被假设为每个专家的清晰数字。然而，知识更新的速度往往快于人们的认知速度，要求专家提供明确的初步意见并确定每个专家的单位调整成本是困难和不切实际的。本文提出了一种新的共识方法来解决上述问题。首先，基于区间值初始意见给出了一种新的距离测度，它保留了专家的初始判断并与大多数实际决策问题保持一致。其次，给出了线性解析公式，以减少分段函数的计算成本。第三，考虑到鲁棒优化在不确定优化中的优势，建立了三种鲁棒成本共识模型来处理共识达成过程中的不确定成本问题。最后，将所提出的方法应用于P2P贷款共识，并进行敏感性分析和比较分析。