知识背景对归纳推理的影响

归纳推理研究领域出现了一些引人注目的实验研究：Heit和Rubinstein等的研究表明了知识背景对归纳推理存在不可忽视的影响；Medin等以“相关”，Loman等以“一致性”作为变量的研究，比较系统地探讨了知识背景对归纳推理的影响；对于这些实验研究结果，假设评价模型、贝叶斯模型、PPP模型以及相关理论都分别能够做出部分的解释；在对上述理论模型进行评价的基础上，提出了利用贝叶斯网络解释知识背景对归纳推理的影响的观点。     

扰动模糊集相似度限定推理方法

作为模糊控制理论基础的模糊推理，近年来得到广泛的研究和应用。结合模糊推理的Mamdani算法与J．EBaldwin的真值限定推理方法，在扰动模糊集上定义了一种模糊推理的新方法一扰动模糊集相似度限定推理(FMP和FMT)方法，提出了扰动模糊集相似度限定推理方法，该方法是满足关系再现的并对关系再现问题进行了讨论，给出了满足还原算法的条件。     

Incorporating Fuzzy Inference in Active Database Rules

Active databases react to stimulation, or event from inside or outside the system without user or application interference through Events Conditions Actions (ECA) rules (triggers). ECA rule is defined as: ON event IF condition THEN action, which means when an event happens, if the condition is satisfied then the corresponding action is executed. The nature of ECA rule makes it an appropriate means to model dynamic character of systems, as gained much studies during recent years. Traditional ECA rule is crisp, which means their events, condition(s) and action(s) are accurate. As indicate that ECA rules can only represent precise knowledge. But knowledge is usually fuzzy in engineering. A concept of fuzzy ECA rules characterized with fuzzy event, fuzzy condition and fuzzy action is proposed in this article. The realization avenues of fuzzy triggers are discussed. The work we have done blazes a way in representing approximate syntax in active database application systems. At last a case of " overheating al     

Solution of a sound field generated by the vibrating surface of a body moving at small Mach number

An approach to calculate the sound field generated by the Vibrating surface of a body with arbitrary shape moving at a small mach number has been developed by using acoustical analogy. A governing equation to calculate the sound pressure on the surface exited by its vibration has been derived on the basis of the united aerodynamics and aeroacoustics approach by Farassat and Long, and the sound field would be determinate. The cases relating to the sound sources of static sphere and moving compact pulsating sphere give a good check for the approach.     

复杂过程建模的模糊集和神经网络方法

为了建立复杂生产过程的被仿真对象的模型,将系统的定性信息（语义规则）和定量信息相结合,提出了一种基于ｐ－范数的近似推理神经网络模型,并给出了一种将梯度方法和最小二乘方法结合起来的交互式学习算法,最后通过实例对算法进行了验证,取得了满意的仿真结果,为进一步科学决策提供了依据     

井下监控系统运行可靠性的一种监测方法研究

研究井下监控系统的可靠性运行状态的监测问题,给出了系统的完整性及完整度的定义,提出并研究基于多传感器信息融合技术的系统完整性监测与估计方法,建立了基于模糊推理方法的系统完整性信息融合监测模型,对于包括众多的,且要求高可靠性的井下分站的监控系统来说尤其重要。     

物元演绎推理

从处理不相容问题的实际需要出发,提出了有关物元演算的若干物元联结词,并建立了一套旨在分析物元之间以及物元要素之间的逻辑关系的物元演绎推理规则,它尤其适用于不相容问题的分析处理。     

联想类比空间收敛理论及同态理论研究

给出了联想类比空间的收敛性及同态结构的一些重要结论，从而进一步丰富和完善了联想类比空间理论；基于此，又给出了一个实现系统ＤＦＧＣＳ     

证据冲突下两种度量模型的一致性条件

讨论了证据冲突情况下的确信度量和概率度量的一致性问题。结论表明：当假设题的先验概率与Ｄｅｍｐｓｔｅｒ组合规则中的归一化常数之间满足一定的关系时，可命名这两种度量方法取得一致的推理结果。此外，还讨论了如何把先验知识引入带有冲突证据的基本推理模型，并给出相应的结论。     

高炉热状态预测专家系统的设计及实现方法

利用人工智能、知识工程及模糊数学思想，开发了高炉炉热预测专家系统．该系统知识表达采用了产生式规则的方式，推理机采用了模糊数学中可信度的方法进行模糊推理，取得了比较满意的效果．