Abstract: An evaluation method based on the combination weighting method by approaching ideal points and unascertained measure model is proposed to evaluate the health status of aero-engines. First, three kinds of weights are respectively calculated by G1 method, Criteria Importance Though Intercrieria Correlation (CRITIC) method and indicator difficulty ratio weighting method. Under the condition that combined weights of indicators should be consistent with the importance of themselves, a programming model is presented to minimize the sum of distances from normal samples to the positive ideal point and abnormal samples to the negative ideal point. Thus, combined weights which could increase the difference between normal and abnormal samples are gotten by the programming model. Then, in order to increase the difference further, every indicator is divided into two opposite classes by K-means algorithm, and each class consists of two small complementary classes in the unascertained measure model. After that, unascertained measures that each indicator belongs to four small classes are calculated separately by the measure function. According to unascertained measures and combined weights, comprehensive measures of samples are obtained. As a result, the health scores of samples are determined by comprehensive measures and the score rule. Finally, the validity of the proposed method is verified by a case study of some military aero-engine and comparisons with other methods.

Key words: aero-engines, health status evaluation, combination weighting method, unascertained measure model

摘要： 为了有效评估航空发动机的健康状况,提出一种逼近理想点的组合赋权法和未确知测度模型相结合的评估方法。首先,分别利用序关系分析法、指标相关性赋权法和指标难度赋权法得到各指标的主客观单一权重;利用逼近理想点的组合赋权法并考虑指标权重与重要性的一致性,求解指标组合权重,使赋权结果更具科学性,加大不同类别评价对象的区分度。其次,基于未确知测度模型,利用K-means算法将各指标分为两个对立等级,每个等级又分为两个互补小等级,以减少信息的重叠进一步区分不同类别的评价对象;根据指标属于不同等级的未确知测度、指标组合权重和评分准则,得到评价对象的健康评分。最后,通过航空发动机实例分析以及与其他方法的对比分析,验证此方法的有效性。

关键词: 航空发动机, 健康状况评估, 组合赋权法, 未确知测度模型