基于一致性局部調(diào)整算法和DEA的語(yǔ)言偏好決策模型

摘要： 設(shè)計(jì)一致性調(diào)整算法和計(jì)算可靠的方案排序權(quán)重向量已成為近年來(lái)備受關(guān)注的兩個(gè)重要研究課題。針對(duì)決策信息為語(yǔ)言偏好關(guān)系的決策問(wèn)題,提出一種基于一致性局部調(diào)整算法和數(shù)據(jù)包絡(luò)分析(DEA)方法的語(yǔ)言偏好決策模型。首先,基于局部調(diào)整策略設(shè)計(jì)一種收斂的乘性一致性調(diào)整算法,該算法不僅使得調(diào)整后的語(yǔ)言偏好關(guān)系具有滿意乘性一致性,而且能夠盡可能多的保存原始決策信息;其次,基于提出的新型語(yǔ)言DEA模型構(gòu)造一種語(yǔ)言偏好決策模型用以確定方案的排序權(quán)重向量,進(jìn)而得到合理可靠的決策結(jié)果。最后,將提出的語(yǔ)言偏好決策模型用于供應(yīng)商的選擇實(shí)例,對(duì)比分析實(shí)驗(yàn)驗(yàn)證了模型的合理性和有效性。
The linguistic preference relations(LPRs) are useful tool to address situations in which the decision makers(DMs) are more comfortable providing their evaluation information linguistically rather than numerical values. Consistency improvement process and deriving the reliable priority weight vector for alternatives are two significant and challenging issues in decision-making with LPRs. In this paper, a linguistic preference decision-making model is investigated with consistency local adjustment algorithm and data envelopment analysis(DEA). Firstly, the concept of multiplicative consistent LPRs is reviewed. Then, a construction approach of multiplicative consistent LPRs is proposed, and a convergent consistency-improving algorithm for LPRs is designed to transform the unacceptable multiplicative consistent LPRs into the acceptable ones, in which the local adjustment strategy is presented to preserve the DM's original decision-making information as much as possible and use the DM's original information sufficiently. Subsequently, a novel linguistic DEA model is developed to explore the relationship between the relative efficiency scores and the priority weight vector of multiplicative consistent LPRs. Furthermore, the linguistic preference decision-making model is investigated to generate the reliable ranking of the alternatives. Finally, a numerical example of selecting the desirable supplier is provided, and the comparison with existing approaches is performed to validate the rationality and effectiveness of the proposed linguistic preference decision-making model.