模糊神经网络在专家系统中的应用研究

为了解决传统故障诊断专家系统知识获取和经验性知识的不确定性的"瓶颈"问题,将模糊神经网络技术引入到专家系统中,介绍了基于模糊神经网络的诊断专家系统的原理、方法、结构及其实现技术.文中将基于模糊多层感知器构成的专家系统应用于柴油机燃油系统故障的分类诊断中,收到了良好的效果.     

基于模糊Petri网的摊铺机液压调平系统故障诊断

以SP120摊铺机的液压调平系统作为研究对象,分析该系统的故障形式及其原因,提出基于模糊Petri网的摊铺机液压调平系统的故障诊断方法,用模糊Petri网表示液压调平系统故障诊断知识,建立模糊Petri网故障诊断推理模型,并利用基于图形搜索的模糊Petri网推理算法对其进行故障的推理和诊断,验证该算法的可行性,该方法应用于诊断工程机械液压系统故障是准确有效的。     

浅谈液压系统故障诊断方法

介绍了液压系统故障诊断的原则及故障诊断常用的几种方法。     

专家系统故障诊断法及其在汽车故障诊断中应用

探讨了专家系统故障诊断法的特点,故障诊断专家系统的基本结构、分类及其在汽车故障诊断中的应用。探讨了故障诊断专家系统存在的不足和解决途径。     

基于Sugeno模糊模型的CNC机床故障诊断方法

通过将Sugeno模糊模型用于数控机床故障诊断系统的方式,实现了对各个诊断专家诊断结果的模糊综合,消除了诊断专家诊断结论的随意性,保证诊断结论的准确.利用此方法基于诊断专家系统技术可以实现对数控机床的自动诊断,诊断结果准确可靠.     

模糊神经网络专家系统在故障诊断中的应用

针对汽车电控发动机故障多,具有模糊性和复杂性的事实,结合模糊逻辑、神经网络、专家系统各自的特点,建立了基于模糊神经网络专家系统的发动机的故障诊断系统;采用面向对象的Visual Basic 6.0、MATLAB 7.0和Ac-cess数据库等软件来实现系统结构,对系统的整个流程,尤其是关键点和重难点进行了研究。结果表明:该诊断系统能对电控发动机故障进行快速准确诊断。     

基于人工智能专家系统的船舶电力系统故障诊断研究

为了提高船舶电力系统在发生故障后锁定故障点的效率,为船员抢修争取到宝贵时间,提高船舶在航行过程中的安全系数.本论文在阐述了船舶电力系统故障特点后,根据故障发生部位的优先级,将诊断模块分为输电线路和其它设备级两个模块,利用matlab程序设计语言设计改进遗传禁忌混合算法预处理数据,并分别使用正向推理和混合推理机制分析数据,同时引入模糊规则,对分析专家系统故障诊断机理在船舶电力系统故障诊断中的改进方案进行了研究,以便于工作人员及时排除故障,对于加强船舶电力系统的稳定性具有重要意义.     

基于规则的机车柴油机故障诊断专家系统

简要介绍了机车柴油机故障诊断专家系统的国内外研究现状，阐述了所研究开发的柴油机故障诊断专家系统的整体框架、诊断知识的表达方法以及诊断的推理机制，并给出了应用专家系统进行机车柴油机故障诊断的实际例子．     

基于模糊灰色关联模型的舰船液压装置故障诊断研究

从舰船液压装置及其故障的特殊性出发,提出液压系统是个灰色系统,其故障具有灰色性质,从而为灰色系统理论应用于液压装置的故障诊断提供了前提.鉴于液压系统的状态特征参数不确定或存在区间取值,基于模糊隶属函数提出了一种模糊灰色关联诊断模型,并对舰船液压装置进行故障诊断研究.结果表明,在信息量较少的情况下,模糊灰色关联分析方法是舰船液压装置状态识别与故障诊断的一个有效工具.     

TY180推土机工作装置液压系统常见故障及诊断

分析了TY180推土机工作装置液压系统的工作原理、组成结构及其典型故障,总结了常见故障的现象及其形成原因,提出了相应的排除措施;运用因果分析法、故障树分析法等故障诊断方法,绘制了推土铲刀不动或上升无力故障的鱼骨图和松土器油缸不动故障的故障树,明确了这2种故障的成因与相应结果之间的逻辑关系。实践证明:运用上述方法能够较快地诊断和排除推土机工作装置液压系统的故障,提高机械工作的可靠性,为工程机械液压传动系统故障检测与维修提供一定的理论依据。     

基于单片机和Fuzzy技术的液压在线故障监测系统的设计

介绍了基于单片机和Fuzzy技术的液压在线故障监测诊断系统的设计,给出了系统组成和软件设计。该系统以高性能单片机为核心,采用Fuzzy逻辑分析技术实现了液压系统故障诊断的基本结构和知识库的方法,建立了液压系统故障Fuzzy分析的逻辑结构,实验表明,该系统提高了液压系统故障推理的正确性和推理速度。     

Fault Diagnosis and Reliability Analysis Using Fuzzy Logic Method

A new fuzzy logic fault diagnosis method is proposed. In this method, fuzzy equations are employed to estimate the component state of a system based on the measured system performance and the relationship between component state and system performance which is called as “performance-parameter” knowledge base and constructed by expert. Compared with the traditional fault diagnosis method, this fuzzy logic method can use human＇s intuitive knowledge and dose not need a precise mapping between system performance and component state. Simulation proves its effectiveness in fault diagnosis. Then, the reliability analysis is performed based on the fuzzy logic method.     

STUDY ON NATURAL LANGUAGE INTERFACE OF NETWORK FAULT DIAGNOSIS EXPERT SYSTEM

Since Internet was born,the computer net workhas developed at a highspeedinits structure and ap-plication.Facing the huge and complex computernet work,we think it hard to deal with the variousfaults of the net work with manual method.Alongwiththe development of AI,the latest achievementof expert system(ES)has provided a new way forthe fault diagnosis and analysis of net work.Butthese ES don't provide a good interface for user.The interface of ES al ways uses command,menuand windows at pres…     

Traditional Chinese medicine diagnosis and evaluation of the clinical effect in post-hepatitic cirrhosis based on fuzzy analysis

In view of the problem of traditional Chinese medicine (TCM) diagnosis, fuzzy comprehensive evaluation is introduced to the diagnosis and evaluation of the clinical effect of hepatitis. A new diagnosis and evaluation model of the TCM clinical effect standard was presented. The old disputed evaluation model by subjective analysis can be displaced by this objective method in its evaluation system. The foundation of this method is the fuzzy diagnosis model. This method has been realized to diagnose hepatitis and its values in practice has been validated through examples of clinical syndromes of post-hepatitic cirrhosis.     

不正常航班恢复策略评估体系研究

基于航班不正常性,从运行控制人员的角度,紧密结合实际运行控制环境,分析不正常航班的恢复策略评估指标体系,建立相应层次结构。鉴于各项指标间的相关性,采用模糊综合评判理论与专家调查法,建立不正常航班恢复策略评估模型,实现对运行控制人员恢复策略的有效评估。     

A Fuzzy Fault Diagnosis Method for Large Radar Based on Directed Graph Model

To meet the requirement of the real-time, accuracy and multi-target diagnosis of the large radar system,a new fuzzy fault diagnosis method based on directed graph model is proposed in this paper. In this method, the large complex system model is defined using the directed graph model firstly, in which the nodes observing the fault by the hierarchical reconstruction of the directed graph are located, then the fault dependency matrix between these nodes and the fault sources are established. And then, we utilize the sensors’ alarm probabilities under different situations to build the characteristic fault observation matrix in the fault observation space. Finally,the optimized corresponding diagnosis method using a fuzzy function, which describes the similarity between the actual observation vector and the fault’s characteristic vector, is designed. The experimental results demonstrate that the proposed method can achieve high diagnosis efficiency and accuracy. It can be widely used in the real radar system.     

故障诊断专家系统应用研究

提出了直流斩波电路故障诊断专家系统的理论根据及设计方法。知识库设计为多元故障树，诊断数据来源于对被诊断对象的测试与知识推理，依据特征向量实现故障定位。采用功能级故障诊断模型比依据网络理论建立的数学模型更为简便。