中文摘要：

为了解决传统故障模式与影响分析（FMEA）的风险评估过程中专家评判的模糊性和不确定性问题,提出基于依赖型语言有序加权几何（DLOWG）算子的FMEA风险评估方法.该方法将故障模式风险优先数（RPN）的各个因子视为语言变量,基于提出的DLOWG算子来综合专家的群体评判意见.该算子对错误或有偏见的语言评价值赋以较低的权重来减轻不公平性对决策结果的影响.基于语言加权几何（LWG）算子和DLOWG算子,考虑各个风险因子及专家意见的权重,提出语言风险优先数的计算公式,对模糊环境中故障模式的风险进行评估.以某汽车外部照明设备的故障模式风险评估为例,验证了该方法的有效性和实用性.

英文摘要：

A new risk evaluation method based on the dependent linguistic ordered weighted geometric（DLOWG） operator was proposed in order to deal with inherent fuzziness and uncertainty of experts＇ judgments in risk evaluation in traditional failure mode and effects analysis（FMEA）.The factors of risk priority number（RPN） of failure mode were treated as linguistic variables.Then linguistic judgments of a team of FMEA experts were aggregated by a developed DLOWG operator,in which the influence of unfair linguistic judgments on the aggregated result can be relieved by assigning low weights to those ＇false＇ or ＇biased＇ ones.Linguistic risk priority numbers（LRPNs）,which considered the relative importance weights of risk factors and expert judgments,were proposed for prioritizing failure modes in fuzzy environment based on the linguistic weighted geometric（LWG） and DLOWG operators.A case of risk evaluation of a car lamp was provided to illustrate the effectiveness and practicality of the method.