基于云模型的信任评估研究

讨论了信任关系的随机性和模糊性共存以及相互融合问题。分析云模型描述不确定性概念的方法和实现定性语意与定量数值相互转换的算法,提出了基于云理论的信任评估模型—信任云。该模型提出云特征参数表达的信任传递和合并算法,在精确描述信任期望值的同时,通过熵和超熵刻画了信任的不确定性。相对于传统的信任评估策略,该模型获取的信任值包含更多的语意信息,更适合作为信任决策的依据。     

正态云模型的重尾性质证明

正态分布和重尾分布在概率研究中具有非常重要的地位,二者具有完全不同的数学形式和物理意义。正态分布的密度函数以指数函数衰减至0,服从正态分布的随机变量,其绝大多数取值在其期望附近,偏离期望很大的取值很少。而服从重尾分布的随机变量,其尾分布函数具有重尾特性,密度函数以幂指数衰减至0。笔者证明了正态云模型是具有均值的重尾分布,是介于正态分布与重尾分布之间的中间状态,正态云模型的参数超熵He是可以实现正态分布向重尾分布转换的桥梁。     

识表示中的不确定性

知识表示一直是人工智能研究中的一个瓶颈,其难点在于知识中隐含有不确定性,即模糊性和随机性.文章提出用云模型3个数字特征(期望值,熵,超熵)来描述一个定性概念,用熵来关联模糊性和随机性.代表定性概念的云的某一次定量值,被称为云滴,可以用它对此概念的贡献度来衡量,许许多多云滴构成云,实现定性和定量之间的随时转换,反映了知识表示中的不确定性.论文以此对我国农历24个节气进行了新的量化解释.云方法已经用于数据开采、智能控制、跳频电台和大系统效能评估中,取得明显的效果.     

基于物元分析与云模型的地下工程围岩稳定性评价

针对以往地下工程岩体质量评判过程中存在的模糊性与随机性问题,根据岩石质量指标、岩石单轴饱和抗压强度、岩体完整性系数、结构面强度系数和地下渗水量这5项指标建立地下工程围岩稳定性评价指标体系.基于物元分析法,通过计算专家评判矩阵的效度物元及合成权重物元来确定评价指标的重要性.根据确定的评价标准,得到各评价因子隶属于各级别的综合云模型,并针对广州抽水蓄能电站二期地下工程岩体实测数据,先通过正向云发生器计算各评价因子所对应的各个评价级别的隶属度,再与其权重结合得到合成隶属度,然后以最大隶属度为依据判别出岩体稳定性所属级别.将该方法所得评价结果与约简概念格-模糊优选评判法、粗糙集-逼近理想解的排序法(RS-TOPSIS)进行对比,三者判别等级一致,表明该方法准确可靠,且免除了对大量样本的学习过程,能够克服评价指标定性语言和定量数值相互转换所产生的模糊随机性,可以用于地下工程岩体稳定性的评判.     

基于DEMATEL-SEWM云模型的露天矿爆破效果评价方法及应用

为了更好处理爆破效果评价中语言模糊性和评价等级定量分析随机性,建立了基于DEMATEL-SEWM的露天矿爆破效果云模型评价方法。运用DEMATEL（决策实验室分析法）和SEWM（结构熵权法）,计算爆破效果评价指标的权重系数,客观反映了指标之间相互关联和指标权重判断差异程度对指标权重的影响。利用云模型特征值E_x、E_n和H_e,客观地描述了爆破效果评价过程中随机性和模糊性。考虑各类指标权重的影响,分析爆破效果综合隶属程度,依据最大隶属度原则确定露天矿爆破效果的评价等级,较好地实现爆破效果评价中定性概念模糊随机性的定量直观表征。运用该模型对某露天矿爆破效果进行了评价,结果表明：各个指标的混合权重分别为0.101、0.102、0.154、0.163、0.128、0.112、0.132、0.107,由权重的大小可知,炸药单耗、大块率和振动速度三个指标对爆破效果影响较大;炸药单耗的原因度为1.2961,其他指标的原因度均小于零,炸药单耗为原因要素,是改善爆破效果的关键要素;四次爆破效果的综合隶属度分别为0.358、0.339、0.36...     

知识表示中的不确定性

知识表示一直是人工智能研究中的一个瓶颈,其难点在于知识中隐含有不确定性,即模糊性和随机性。意提出用云模型3个数字特征（期望值,熵,超熵）来描述一个定性概念,用熵来关联模糊性和随机性。代表定性概念的云的某一次定量值,被称为云滴,要以用它对此概念的贡献度来衡量,许许多多云滴构成云,实现实性和定量之间的随时转换,反映了知识表示中的不确定性。论以此对我国农历24个节气进行了新的最化解释。云方法已经用于数据开采、智能控制、跳频电台和大系统效能评估中,取得明显的效果。     

三级倒立摆的云控制方法及动平衡模式

文章提出了定性和定量之间转换的云模型的形式化表示方法,用来反映语言值中蕴涵的模糊性和随机性,给出云发生器的生成算法,解释多条定性推理规则同时被激活时的不确定性推理机制。利用这种智能控制方法有效地实现了单电机控制的一、二、三级倒立摆的多种不同动平衡姿态,显示其鲁棒性,并给出了详细试验结果。研究成果不仅可用于对太空飞行器以及机器人控制,而且对揭示定性定量转换规律和策略具有普遍意义。     

模糊变量与随机变量组合时模糊可靠度计算方法研究

对现有的随机应力、模糊强度可靠度计算方法进行了系统的分析研究,将其分成模糊变量直接转化为随机变量、由模糊强度的隶属函数直接构造模糊失效事件的隶属函数、可能度法和实用计算法四类,根据模糊失效概率的属性又可分成定量计算法和模糊计算法两类;得到了在定量计算方法中,广义密度函数法与面积法等价,当量密度函数法与截集法等价,真正独立的主要方法为广义密度函数法、当量密度函数法和直接转化法三种等结果;对强度模糊隶属函数为线性、正态分布情况推导了有关公式;通过算例给出了各主要方法的计算结果,其值相差较大,因此,哪种方法更合理还必须作进一步研究。     

模糊分布参数的全局灵敏度分析新方法

为合理度量随机输入变量分布参数的模糊性对输出性能统计特征的影响,提出了模糊分布参数的全局灵敏度效应指标,并研究了指标的高效求解方法。首先,分析了不确定性从模糊分布参数至模型输出响应统计特征的传递机理,以输出性能期望响应为例,利用输出均值的无条件隶属函数与给定模糊分布参数取值条件下的隶属函数的平均差异来度量模糊分布参数的影响,建立了模糊分布参数的全局灵敏度效应指标。其次,为减少所提指标的计算成本、提高计算效率,采用了扩展蒙特卡罗模拟法(EMCS)来估算输入变量分布参数与模型输出响应统计特征的函数关系。最后通过对算例的计算,验证该文所提方法的准确性和高效性。     

基于熵技术的矿山安全模糊综合评价

根据矿山安全评价所存在的模糊性和随机性的特点,在建立多级模糊综合评价的指标体系的基础上,借用信息论中熵的概念计算客观权重的方法,运用模糊数学理论,建立二级模糊综合评价模型,提高了矿山安全现状评价的客观性和科学性,具有一定的工程实际应用价值.     

A mathematical programming approach to supply chain models with fuzzy parameters

Supply chain (SC) models play an important role in supply chain management (SCM) for reducing costs and finding better ways to create and deliver value to customers. An approach to deriving the membership function of the fuzzy minimum total cost of the multi-product, multi-echelon, and multi-period SC model with fuzzy parameters is proposed in this article. On the basis of α-cut representation and the extension principle, a pair of mathematical programs are formulated to calculate the lower and upper bounds of the fuzzy minimum total cost at possibility level α. The membership function of the fuzzy minimum total cost is constructed by enumerating different values of α. To demonstrate the validity of the proposed procedure, a four-echelon five-period SC model with fuzzy parameters is solved successfully. Since the objective value is expressed by membership functions rather than by crisp values, they completely conserve the fuzziness of input information when some of the SC data are ambiguous. Thus the proposed approach can represent SCs with fuzzy parameters more accurately, and more information is provided for designing SCs in real-world applications.     

Incorporating fuzzy weather-related outages in transmission system reliability assessment

Fuzzy models for weather-related outages of overhead lines and a combined probabilistic and fuzzy technique for transmission system reliability assessment are presented. The region-divided weather states are modelled using a probability approach. This approach and the fuzzy models of weather-related outages are combined and incorporated into a Monte Carlo simulation procedure of transmission system reliability assessment to evaluate membership functions and mean values of reliability indices. The reliability test system is used to demonstrate an application of the proposed models and method. The membership functions of reliability indices provide a wider insight into the fuzziness of weather effects, which cannot be modelled by traditional probability concepts.     

Optimal echelon stock policies for multi-stage supply chains in fuzzy environments

This paper investigates decision-making in multi-echelon serial supply chain management in the presence of imprecision or uncertainty arising from human reasoning, emphasising the computational resolution. The proposed analysis method is based on a combination of the extension principle and the alpha-representation in fuzzy theory and optimisation theory. The problem is first formulated as a fuzzy optimisation model with several fuzzy parameters. To conserve the fuzziness of the input information of the supply chain, such as forecast market demands and inventory costs, a pair of two-level mathematical programs is proposed to identify the lower and upper bounds of the fuzzy performance at different possibility levels, so that the complete membership function can be described. Four example scenarios are solved to demonstrate the validity of the proposed analysis method. The proposed methodology is widely applicable with different types of membership functions for fuzzy parameters, positive lead times or other more complicated cases. The managerial implications are also discussed for reference by decision-makers.     

含非闭合隶属函数模糊变量的结构失效概率分布研究

结合工程实际,提出了非闭合隶属函数的截断可能性分布模型,并对模糊强度和模糊应力进行截断处理,给出了结构模糊随机失效概率随截断参数的分布,并给出了结构模糊随机失效概率分布的数值计算方法。所提出的方法不仅可以考虑基本变量的随机模糊性,而且可以考虑安全和失效状态的随机模糊性。关于强度和应力两个基本变量的情况易于推广应用到多个变量的情况,以解决多变量体系中含有非闭合隶属函数模糊变量的安全分析问题。     

Appearance of the Sun and the Moon seen through clouds

If the Sun can be seen at all through thin clouds it usually has a sharp edge, although occasionally it appears fuzzy, especially through altostratus, but rarely, if ever, through fog. Experiments with suspensions of polystyrene spheres of different sizes and optical thicknesses suggest that the range of cloud optical thicknesses over which a fuzzy Sun is seen increases with particle size. Nonsphericity, turbulence, and cloud horizontal inhomogeneity are not necessary for fuzziness. A possible explanation for what is observed is that, for a given optical thickness, the modulation contrast function of a cloud decreases more rapidly with increasing frequency the greater the particle size. Consequently the transition from optical thicknesses for which contrast is above the contrast threshold at all spatial frequencies to optical thicknesses for which contrast is below the threshold at high frequencies is sufficiently gradual to permit fuzziness of the Sun to be observed through clouds of constantly changing opt cal thickness.     

Non-linear programming for the optimization of machine repair problems in fuzzy environments

Machine repair models have wide applications in many systems such as production line systems and maintenance operations. A procedure is developed to derive the fuzzy objective value of the cost-based machine repair optimization problem, in that the cost coefficients and the machine breakdown rate are fuzzy numbers. On the basis of the extension principle, a pair of non-linear programs are formulated to calculate the lower and upper bounds of the fuzzy minimum expected total cost at the possibility level α. The membership function of the minimum expected total cost is constructed by enumerating different values of α. A numerical example is solved successfully to demonstrate the validity of the proposed method. Since the minimum expected total cost is completely expressed by a membership function rather than by a crisp value, the fuzziness of the input data is conserved, and more information is provided for decision-making. Furthermore, since the optimum repair rate obtained is fuzzy, a crisp optimum repair rate based on the Yager ranking indices is recommended for practical use.     

基于模糊规划的间歇过程生产调度建模

通过对现有生产调度模糊模型的分析，针对间歇生产过程，提出了一种新颖的参数模糊的通用模糊规划建模方法，模糊模型采用2种基于普通遗传算法的模糊算法，即模糊模拟和SFA算法。该方法中，参数隶属函数选取灵活，采用合适的模糊表示方式。最后通过实例验证了该模糊建模方法是一种有效的方法。