自经典的Hegselmann‑Krause连续观点动力学模型提出以来，许多学者对其进行了大量的研究和改进，但如何具体的在观点更新过程中融合外部群体意见的影响尚未被考虑其中。本文基于心理学中的乐队花车效应，考虑群体意见和信任程度等因素，提出了一种新的基于乐队花车效应的Hegselmann‑Krause观点动力学模型。该模型中，意见领袖与普通个体对来自其相邻节点意见的信任系数被建模为当前时刻个体意见与群体意见之差的函数，并在此基础上提出了的意见更新规则。通过在人工生成网络和真实网络中进行实证分析表明，在群体意见和意见领袖的共同影响下，无论意见领袖比例、个体置信度和网络规模如何变化，除固执个体和孤立个体外，其他个体的意见都能在较短时间内达成一致，从而为社交网络中人们观点的从众现象提供了新的解释。

Many researchers have made various improvements to the Hegselmann‑Krause model after it was proposed. However, no one has considered the influence of group opinion during the opinion update process. In this paper, a novel bandwagon effect‑based Hegselmann‑Krause opinion dynamics model is proposed considering factors such as group opinion and trust degree. In our work, the trust degree of opinion leaders and normal individuals to their neighbours is formulated to a special function of the difference between their and their neighbours′ opinions. Then, the update rules of the agents′ opinions in the social network are developed. Numerous experiments conducted on both artificially generated network datasets and real‑traced network datasets show that under the joint influence of group opinion and opinion leaders, the opinions of all the agents, except the stubborn and isolated ones, can reach a consensus in a short time, whatever the proportion of opinion leaders, the confidence bound or the network scale is, which can well explain the conformity phenomenon in social networks.