基于GSM-熵值法和TOPSIS排序的系统风险评估模型

以汽车零部件作为风险评估的对象,基于系统工程的理论,将汽车零部件划分出多个子系统。工艺质量方面,提出了一种利用高斯模型对子系统的特征参数进行故障风险诊断的方法。标准化生产的零部件可以通过采集前样本,然后用特征向量来建立高斯模型。引入对数似然概率和反熵法对各个子系统的风险进行排序。故障影响方面,提出了一种新的系统风险评估的方法。以FMEA的严重度、发生度和被检测难易度作为风险指标,引入三角模糊函数建立决策矩阵,并对矩阵进行修正,利用TOPSIS法进行评估。最后综合两种评估结果,给出了组合风险排序的算法公式。     

地铁车门故障模式的可靠性评估

车门系统是地铁列车的关键子系统,对其进行可靠性评估从而确定危害性较大的故障模式有助于提高车门的可靠性和改善车门维修决策。由于车门故障数据的不足,难以对风险因子及其权重进行准确评价而导致车门可靠性评估存在模糊性,故引入模糊理论构建模糊TOPSIS方法并应用于车门故障模式的可靠性评估中。对风险因子进行模糊评价得到加权决策矩阵,应用加权决策矩阵计算每一个故障模式到正负理想方案的相对贴近度,得出了EDCU功能失效、螺母组件破损和行程开关S1破损对车门危害性较大的评估结果。分析结果与专家经验一致,为车门的设计与维修决策提供了技术支持。     

一种面向制造企业的供应商优选方法

针对制造企业的零件供应商选择问题,提出一种用模糊粗糙方法来计算指标值,用熵权来计算指标权重的优选方法。将评价指标分为质量、供应、经济、合作和环境5项,构造了一种模糊粗糙方法;用梯形模糊数来表示专家评分,通过粗糙近界区间来计算供应商在评价指标上的取值;采用熵权法来实现评价指标的自适应赋权;最后使用Kullback-Leibler散度替代欧氏距离的改进逼近理想解排序(Technique for Order Preference by Similarity to an Ideal Solution,TOPSIS)来对供应商的加权指标值向量进行排序。通过一个企业应用实例验证了该方法的正确性和可行性。     

基于TOPSIS法的数控车床刀架系统故障分析

基于TOPSIS理论从故障发生频率、维修时间、维修费用和对相关部件的影响程度4个方面对数控车床刀架系统进行故障分析,通过建立决策矩阵,计算理想解和负理想解,得出刀架系统各故障模式的相对贴近度,最终得到各故障模式重要度的排序,通过与故障模式、影响及危害性分析的评价结果进行对比,得出应用TOPSIS方法的刀架系统故障模式排序更合理。     

基于模糊决策的电力变压器风险评估方法

针对传统风险评估方法存在评价指标等同对待、当风险优先级(RPN)相等时无法进行评估、评估结果准确性低等不足,提出了基于模糊多准则决策的电力变压器风险评估方法,并成功应用于电力变压器故障模式的风险评估.该方法在对电力变压器进行故障模式及影响分析(FMEA)的基础上,选择发生度、严重度和可检测度作为电力变压器的风险评价指标;采用三角模糊数对3项风险评价指标分别进行模糊化处理,运用熵重法确定各评价指标的权重以表现其相对重要度;通过使用模糊多准则决策理论对故障模式进行排序最终得到风险评估结果.评估实例表明,该方法不仅有效地克服了传统评估方法无法实现评价指标的区别对待和当风险优先级(RPN)相等时无法进行评估这两大不足,而且能有效提高评估结果的准确性.     

基于三角模糊数TOPSIS法的数控机床危害性分析

为提高数控机床经典危害性分析的准确度,提出了基于三角模糊数与TOPSIS法的数控机床危害性分析方法。综合考虑故障的频率、故障对系统的影响、探测度、停机时间及维修的难易程度等影响因素,建立数控机床的三角模糊数评价决策矩阵,并基于TOPSIS法,计算子系统各评价指标与正理想解和负理想解的相对贴近度,给出各子系统的危害性排序。通过对多台XK7132型数控机床的故障现象进行分析,结果表明,主轴系统的危害性最高;与经典危害性分析方法进行对比,文中方法更加科学、合理,为提高数控机床的可靠性水平提供了理论依据。     

基于拓展TOPSIS的风险型产品服务系统设计方案选择方法

针对产品服务系统(Product Service System,PSS)方案评价的风险性和不确定性,提出结合累积前景理论的拓展TOPSIS方法对PSS设计方案进行评价。采用区间直觉模糊数获取并表达专家对候选方案的评价信息;将区间直觉模糊相对熵作为距离的度量,将决策矩阵转化为相对于参照点的损益决策矩阵,基于累积前景理论计算每个方案的指标综合前景值;以正理想前景值和负理想前景值为参考点,采用基于相对熵的TOPSIS方法对候选PSS方案进行排序,最后通过一个算例说明了该方法的可行性和有效性。     

面向绿色再制造的机械产品综合评价与应用

为了科学地对再制造机械产品进行绿色综合评价,考虑到决策者选择时的主观性和模糊性,提出一种将模糊层次分析法与熵权理想点法集成的求解模型。运用模糊层次分析法从时间属性、技术属性、经济属性、资源消耗以及环境影响建立再制造机械产品评价体系,并以MATLAB为计算工具,用模糊层次分析法和熵权法分别对评价指标进行主、客观权重组合赋权,根据改进的理想点法结合各指标主客观权重建立加权规范化决策矩阵并得到各方案贴近度,依据贴近度的大小对再制造产品的优劣进行排序,从而评价出再制造最佳的机械产品,通过工厂实例分析证明该方法的合理性与可行性。     

基于改进TOPSIS法的航空发动机性能评估方法

针对TOPSIS改进方法的逆序和中垂线矛盾问题,本文以原点为负理想解,对航空发动机性能参数进行标准化处理,并提出了基于网络层次法参数权值计算方法,通过以表征参数相关性的马氏距离计算性能贴近度,最后以PW系列8台发动机航后信息作为输入量进行评估验证。结果表明:基于改进TOPSIS的发动机性能评估模型解决了TOPSIS法模型因逆序和中垂线矛盾问题引起评估结果偏差较大的问题,能准确地反映发动机性能状态,同时与ANP法评估结果的一致性说明改进TOPSIS法得到的发动机性能排序是可靠性,为发动机性能评估提供一种新方法。     

模糊随机环境下的汽车租赁方案指标权重确定方法研究

为解决指标权重未知的模糊-随机多属性决策问题,构建了基于相似性差异最小化的权重优化模型确定指标权重,采用逼近理想解法实现决策过程。首先,运用区间直觉模糊集和随机占优理论定量化处理决策者的模糊-随机评价信息;其次,基于初始决策矩阵和综合决策矩阵之间的相似性差异构建指标权重优化模型;最后,根据定义的正、负理想解,运用区间直觉模糊交叉熵和欧氏距离分别测定模糊指标和随机指标的偏离度,基于改进的TOPSIS框架整合模糊指标和随机指标的相对贴近度以实现方案排序。实验显示,所提方法更能够方便地确定各指标权重。分析表明,所提方法避免了其它权重确定方法的不足之处,具有很好的敏感性,适用于指标权重未知的模糊-随机多属性决策问题。     

基于粗糙集-田口质量观的失效模式与影响分析

针对目前失效模式与影响分析（FMEA）方法中对实际风险较大的失效模式的风险难以有效评估的问题,提出了一种基于粗糙集理论和质量损失函数的FMEA新方法。首先,在对潜在失效模式进行分析的基础上,通过专家系统构建失效模式风险评估矩阵模型;其次,以降低数据本身的主观性和模糊性为目标,基于粗糙集理论将风险评估矩阵模型转换成粗糙域矩阵模型;再次,根据求解的粗糙域矩阵,运用多元质量损失函数模型,对每一个失效模式的潜在损失进行求解,依据潜在损失的大小对失效模式风险进行排序。最后,以发动机装配过程为例,对其失效模式与影响进行分析,并将其分析结果与目前的FMEA方法比较,验证了所提出方法的合理性和有效性。     

粗糙云模型与PROMETHEE的航发主轴轴承失效模式重要度评估

针对目前航空发动机主轴轴承关键失效模式确定困难的问题,提出了一种粗糙云模型与偏好顺序结构评估（PROMETHEE）的航发主轴轴承失效模式重要度评估方法。结合失效模式与影响分析（FMEA）方法,通过专家系统构建主轴轴承各失效模式的风险评估矩阵;运用粗糙集理论与云模型理论,将风险评估矩阵转换为粗糙云评估矩阵,描述专家评价信息中存在的主观性、模糊性和随机性问题;运用PROMETHEE计算各失效模式的流出量、流入量与净流量,根据净流量的大小对航发主轴轴承的6种失效模式进行风险排序,为制定预防措施和降低风险提供可靠依据。     

油田站场输油泵模糊安全评价

输油泵机组作为输油站场的核心动力设备,负责为原油提供压能。现以某输油站场输油泵机组为例进行安全评价,根据故障模式、原因及影响分析对输油泵机组进行部件划分,分析输油泵机组的故障模式,建立输油泵机组的安全评价体系,应用模糊关系矩阵,采用层次分析法确定指标因素权重,最后模糊综合评价输油泵机组,确定出该输油站场6台输油泵机组的安全等级及各指标因素权重值,针对故障影响分析提出改进措施。     

NUMERICAL SIMULATION ALGORITHM FOR RELIABILITY ANALYSIS OF COMPLEX STRUCTURAL SYSTEM BASED ON INTELLIGENT OPTIMIZATION

An efficient importance sampling algorithm is presented to analyze reliability of complex structural system with multiple failure modes and fuzzy-random uncertainties in basic variables and failure modes. In order to improve the sampling efficiency, the simulated annealing algorithm is adopted to optimize the density center of the importance sampling for each failure mode, and results that the more significant contribution the points make to fuzzy failure probability, the higher occurrence possibility the points are sampled. For the system with multiple fuzzy failure modes, a weighted and mixed importance sampling function is constructed. The contribution of each fuzzy failure mode to the system failure probability is represented by the appropriate factors, and the efficiency of sampling is improved furthermore. The variances and the coefficients of variation are derived for the failure probability estimations. Two examples are introduced to illustrate the rationality of the present method. Comparing with the direct Monte-Carlo method, the improved efficiency and the precision of the method are verified by the examples.     

机械设计方案模糊熵权综合评价

针对机械设计方案中存在的诸多模糊因素,提出了一种模糊熵权综合评价方法。该方法首先对机械设计过程中所出现的模糊因素进行分类,并建立相应的评价指标分级矩阵,随后借助信息工程中的熵权理论,利用所获得的分级矩阵中各评价指标的差异程度计算其权重,然后再对分级矩阵进行加权,最后利用信息熵实现对所选设计方案的总体评价。文末给出了实例,验证了该方法的可行性与有效性。     

面向装备健康管理的监测参数选择与健康评估方法研究

针对装备健康管理中监测参数选择缺乏定量方法和健康评估中参数权重确定过于主观的问题,提出了基于模糊层次分析法的监测参数选择方法和基于粗糙集的健康评估方法。该方法在常规层次分析法基础上,引入三角模糊数构造判断矩阵,克服了判断矩阵中元素的不平衡性,符合专家对评判对象判断存在模糊性和不确定性的客观事实。给出了监测参数选择的详细步骤,基于所选监测参数,把健康评估问题描述成典型的多属性决策问题,利用粗糙集属性重要度确定参数客观权重。航空发动机中的应用结果表明,该方法具有较好的有效性和可行性。     

Selecting Plant Location via a Fuzzy TOPSIS Approach

A fuzzy TOPSIS approach for selecting plant location is proposed, where the ratings of various alternative locations under various criteria and the weights of various criteria are assessed in linguistic terms represented by fuzzy numbers. In the proposed method, the ratings and weights assigned by decision makers are averaged and normalised into a comparable scale. The membership function of each normalised weighted rating can be developed by interval arithmetic of fuzzy numbers. To avoid complicated aggregation of fuzzy numbers, these normalised weighted ratings are defuzzified into crisp values. A closeness coefficient is defined to determine the ranking order of alternative locations by calculating the distances to both the ideal and negative-ideal solutions. Using the suggested method, the decision makers’ fuzzy assessments with different rating viewpoints and the trade-off among different criteria are considered in the aggregation procedure to assure more convincing decision making. A numerical example demonstrates the feasibility of the proposed method. ID="A1"Correspondance and offprint requests to: Dr T.-C. Chu, Institute of Industrial Management, Southern Taiwan University of Technology, 1 Nan-Tai Street, Yung Kang City, Tainan County 710, Taiwan. E-mail: mrchu@mail.stut.edu.tw     

[数据驱动的产品设计]复杂产品设计方案的数据驱动多属性优化决策

At present, the quality of product design scheme mainly depended on the experiences and levels of designers, to solve this problem, a data-driven multi-attribute decision-making method for product design schemes was proposed, which might be applied to the evaluation of the best program at all stages of product design. Each group of reviewers used the TOPSIS method to evaluate the scheme, for the weight of decision criteria, the entropy weight method was used to determine the weights, the Borda function method was applied to assemble the decisions of each group, the final decision results were obtained, and the comprehensive evaluations of the design scheme were completed. This method was more fully used for data, avoiding subjectivity to the greatest extent, and the evaluation results were more reasonable. Finally, the method was applied to the design of oilfield polymer injection devices, and the opinions of designers, processors and managers are gathered to provide more comprehensive optimization decision results.     

An integrated fuzzy MCDM approach and analysis to evaluate the financial performance of Iranian cement companies

In today's highly competitive environment, suitable appraisal of a company's performance is of vital importance not only for the company but also for its suppliers and customers. The main objectives of this study were to analyze the financial ratios of Iranian cement-producing companies and to develop a fuzzy model to evaluate financial performance. The proposed fuzzy approach is based on the fuzzy analytical hierarchy process (FAHP) and the technique for order preference by similarity to ideal solution (TOPSIS) method. In contrast to other studies in the literature, in the present study, the FAHP method was used in determining the weightage of criteria by decision makers and the TOPSIS method was used in determining the ranking of the companies. The proposed combined method was used to evaluate the performance of eight Iranian cement companies listed in the Tehran Stock Exchange using data from their financial tables and subsequent ranking of these companies. The final results of the analysis revealed the performance ranking of the companies Sabhan, Sarab, Sedasht, Safar, Sekaroun, Sakarma, Sanir, and Sahrmoz with priority scores of 0/55, 0/51, 0/50, 0/49, 0/42, 0/37, 0/36, and 0/33, respectively. The results indicate an overall performance ranking because the weights of the criteria were determined by decision makers with different experiences, positions, and proficiencies. If it is desired, it is possible to make an evaluation only for creditors, investors, or shareholders; however, in such cases, the weights of the criteria will vary and the ranking of the companies can change. The proposed method can also be applied to the evaluation of companies in other sectors.