<p>In the consensus-reaching process, individual decision-maker behavior, interaction behavior between the moderator and decision-makers, and fairness have been extensively studied. However, these factors are often discussed separately, which may not match practical decision-making environments. In this paper, we propose an improved consensus mechanism with maximum-return and minimum-cost (IMRMCCM) based on the Stackelberg game framework, which integrates deviations in personal utility values, individual tolerance, and compromise limit behaviors. Different from the classical compromise limit, the compromise limit in this paper is defined by the personal utility function of each decision-maker. In particular, it can be exceeded in the consensus-reaching process and, interestingly, the unit compensation cost decreases when the decision-maker makes adjustments beyond the compromise limit under the utility truncation assumption. Furthermore, we propose a limited cost consensus mechanism with maximum-return and maximum-fairness (MRMFCM). One advantage of MRMFCM is that it integrates individual decision-maker behavior, interaction behavior between the moderator and decision-makers, and fairness, which makes it more consistent with practical decision-making environments. Then, we propose a penalty function-based differential evolution algorithm to solve IMRMCCM and MRMFCM. Simulation experiments on green supply chain management demonstrate the effectiveness of the proposed consensus mechanisms.</p>