故障分析

Based on image processing and database technique, the processing flow of image data analysis and fault hidden danger detection by overhead power line mobile inspection system as well as the conservation and management scheme of fault information are proposed.

基于图像处理技术和数据库技术，提出了架空输电线路机器人巡检系统分析图像数据、检测故障隐患的处理流程，以及故障信息的保存管理方案，介绍了图像信息处理软件分析可见光图像和红外图像、获取故障特征值的过程，使用符合行业标准的检测方法对故障特征值进行判断，并形成可视化故障检测报告。

On the basis of dual feature identification method, combined with small wave analysis, it effectively extracts failure features, analyzes the stability power failure vibration signal feature on the general Fourier atlas, analyzes the shock power failure vibration signal feature on the resonance modulating Fourier atlas, using small wave to analyze the shock failure vibration signal feature.

是一种在"双特征判别法"的基础上，结合小波分析，有效提取故障特征，进行诊断的"三特征判别法"，所述的故障信号"三特征判别法"就是：在常规傅立叶谱图上分析平稳功率故障振动信号特征，在共振解调傅立叶谱图上分析冲击功率故障振动信号特征，利用小波分析冲击故障振动信号特征。

Since the 1960s,the leading shipbuilding countries begin to study on the ship condition monitoring and trouble analysis, Nobodys engine-room shipping on duty appeared .certainly the condition monitoring and trouble analysts system also meets some question ,such as: only give alarm or show the actual value of the parameter, does not possess the function of analyzing and giving operator the direct suggestion function about the complicated trouble reason , Only monitor the present running conditions, does not possess the function of predicting equipment future and run reliability though trouble analysis technology makes good progress in these yeas, onboard the ship is still rare .aiming to improve the shipping safety, this paper make some research work on the ship engine-room, the main innovation are as follows:Firstly, this paper introduces and analyzes the research work and current situation on the condition monitoring, trouble analysis and safety assessment .especially in the maritime fields .

船舶机舱系统结构复杂，故障源多，征兆参数丰富，故障与征兆之间常常不存在简单的对应关系，而是相互影响，关系错综复杂。60年代以来，世界上先进的造船国家都相继开展了船舶机械状态监控和故障诊断的研究，出现了无人值班机舱船舶，但仍存在不足之处，如：只给报警提示或显示参数的实际值，不具备对复杂故障的原因分析并给操作者以直接的提示或建议功能；只监测当前的运行状况，不具备预测设备未来运行可靠性的功能；尽管近年来设备故障诊断技术得到了很大的发展，但真正能成功应用于实船的尚不多见。本论文以船舶机舱为对象，以提高船舶营运安全性为目标，来研究船舶机舱的安全控制技术，主要内容如下：首先，介绍和分析了状态监测、故障诊断和安全评估的研究内容和现状，尤其是在船舶领域的研究近况。

Aimed at the characteristic of multiple types of faults possibly happened in nuclear power plant, large scale of training sample, and requirement of quick and accurate diagnosing, after the analysis of large sample number generated by large fault type exceeds limitation of 64K paragraph in DOS and large number of local minmum in error surface, measurements of same format of data file, Win95/NT operating system platform, resetting weight learning rate, dynamic training set in quick learning algorithm and improvement on quick learning algorithm using homotopy method which can avoid local minmum points in error surface have been adopted to ensure quickly and effective process of the course of neural network's training and testing.

针对核动力装置可能发生的故障种类多，训练样本规模大，故障诊断需要快速准确的特点等，分析了样本多和故障种类多产生的训练样本量超出DOS段大小和网络误差曲面上局部极小值多的情况，采取了使用相同格式数据文件、Win95/NT操作系统平台，对快速学习算法采用重置神经网络权值学习率和动态训练集、并采用能够有效克服网络误差曲面上局部极小点的同伦方法对学习算法进行改进等措施保证神经网络训练测试过程的快速、有效地进行；针对装置发生的故障须快速、准确诊断的需要，分析了故障的产生对装置参数变化的影响及操纵员对故障诊断的基础，在核动力装置发生故障时参数曲线的变化量与正常运行时参数曲线的变化量存在明显差异的基础上，提出采用参数曲线的变化量作为神经网络的输入，并围绕参数变化量的方法采用二次曲线拟合滤波求变化量和阈值技术来保证神经网络得到精确的装置参数变化量，从而得到准确的诊断结果。

At the same time, on the base of analysis of the present situation, the development trend and the leading edge technology, the necessity and feasibility about the exploitation of the full hydraulic bulldozer"s sate monitoring and fault diagnosis system is given; According to the structure, working feature and working environment, it works out system schema and system composition; The system is divided into display module, communication module and fault diagnosis module by means of block design, the implementation method and key parameter are also given; In light of the features of the full hydraulic bulldozer, the state monitoring and fault diagnosis platform composed of controller RC6-9 and display DI2/10 is designed based on CAN-Bus and the hardware design is completed; It lists the measured values and state of monitored parameters which should be real-time displayed in monitoring module as well as the data which will be transmitted; Furthermore, On the base of the study of, it gives 34 kinds of constant faults" diagnosis methods; Using BODAS software and DPS software, it compile the program code of the controller and the displayer; At last, this paper design and make the simulation test device of control system of construction machinery, through the experiment, the realizability, the categoricalness, the reliability, the stability and real-time are tested, in addition, the improved methods are given.

本文收集整理了国内外智能推土机研究的相关文献资料，综合分析了国内外智能推土机的研究现状、发展趋势和前沿技术，论证了开发全液压推土机智能系统的重要组成部分&全液压推土机状态监测与故障诊断系统&的必要性和可行性；根据全液压推土机的结构、作业特点和作业环境，制定了全液压推土机状态监测与故障诊断系统方案，确定了系统的结构组成；采用模块化设计方法，将整个系统分成了显示、通信和故障诊断三个模块，给出了各模块功能的实现方法，确定了各模块的关键技术参数：选用RC6-9和DI2／10作为系统的控制器和显示器，用CAN总线搭建了系统的通信平台，完成了系统的硬件设计；确定了系统监测参数和通信数据，对全液压推土机常见故障特征进行了分析，给出了34种常见故障的诊断方法；在BODAS软件和DPS软件的集成编程环境下，编写了控制器和显示器程序；设计并制作了工程机械控制系统模拟试验板，并在此基础上通过多次的程序调试与试验，对所设计的全液压推土机状态监测与故障诊断系统功能的可实现性、完备性、可靠性、稳定性和实时性进行了检验，证实了此系统的可行性，并提出了进一步的改进方案和措施。

Faults of ball milling system in power plant are analyzed by means of fault tree analysis. A fault tree is established that contains faults of ball mill, coal feeder, mill exhauster and separator. The minimal cut set is figured out. The cause for the faults are detailed.

利用故障树分析法对火力发电厂制粉系统的常见故障进行了分析研究，建立了制粉系统的故障树，涵盖了球磨机、给煤机、排粉机、粗细粉分离器的常见故障，求出了其最小割集，给出了制粉系统发生故障的原因及导致故障各种原因组合的可能性。

Firstly, This text has described information fusion technology characteristic, form structure and concrete processing method, and application of information fusion in fault diagnosis proves the feasibility and validity from the angle of information theory. Based on this,it is combined the fault diagnosis of aeroengine with Information fusion, a method and model which the fault diagnosis of aeroengine information fusion is brought forward. a research which combined neural networks with D-S evidence is brought forward. Based on the analysis of single neural network characteristic, the model of roted neural network is put forward. With the simulation based on the example, we can know evidence theory fusion improved the fault diagnosis It is hard to give the basic probability function in the process of using the evidence theory, So the model based on network and D-S evidence is put forword.

本文首先论述了信息融合技术的特点、形式结构和具体的处理方法，并从信息论的角度论证了信息融合技术在故障诊断中的可行性和有效性；在此基础上，将航空发动机故障诊断与信息融合相结合，提出了一种基于信息融合技术的发动机故障诊断模型和方法，并在此基础上，提出和分析了基于人工神经网络和D-S证据理论的信息融合故障诊断，研究了神经网络的建模方法、组建原则和实现策略，并结合诊断实例进行了分析；提出了基于D-S证据理论的决策融合，阐述了D-S证据理论的方法，模型，并结合算例进行了分析。

First, it diagnoses and analyses the maglev train's suspension system and traction system by applying the FTA and fuzzy FTA method. Furthermore, it presents the new way to solute the diagnosis searching arithmetic by building the searching matrix based on the factors such as components' key weightiness, faults happen probability, searching cost. Then it can get optimal searching sequence for seeking the reason to the faults. Second, it builds the train's rank evaluation model based on multiple hiberarchy neural networks, after it takes the strategy that set the component diagnosis and whole train fault rank evaluation into two steps. Third, it deeply analyses reliability of the maglev train's power system with function-correlated feature, by the method that combining the FTA with the Markov model. Finally aiming at maintenance ahead of the fault, it proposes the blue print for the maglev train's maintenance. What's more, it introduces the design and realization to recorder for the train run data in detail.

应用故障树和模糊故障树分析法对磁浮列车悬浮系统、牵引系统进行全面的诊断分析，提出了以部件关键重要度、故障概率、搜索代价为因素，建立搜索决策矩阵，综合排序的搜索算法；采取部件诊断和列车集中评估相结合的分层诊断策略，建立基于多重神经网络的列车故障等级评估模型；结合故障树和马尔可夫模型对具有功能相关特性的列车电源系统进行可靠性分析；着眼事前维修，提出了列车运行维护方案，并设计实现了列车运行数据记录器。

Two GFRF system model identify algorithms based respectively on Signal Flow Graph/Driving Point Impedance and EDA simulation are proposed. In order to solve the problem of fault diagnosis that the inner structure and parameters of the Object under Test cannot be able to determine, a new fault diagnosis method based on the phase space reconstruction and GFRF general formula is investigated. This method builds up a bridge between the chaos theory and Volterra series analysis, which does not strongly depend on the topological structure and the testing data length, is sensitive to the fault property, traces monotonously the change of system fault, and possesses some significant predominance compared with conventional fault diagnosis methods of analog circuits.

提出了基于信号流图/驱动点阻抗法的GFRF算法和基于EDA仿真和非线性频谱响应特性分析的系统模型辨识算法，为解决工程实践中不能确知被测对象内部结构和参数时的故障诊断问题，研究了基于相空间重构和GFRF简化算法的复杂电子系统故障诊断方法，该方法建立了混沌理论与Volterra级数分析法之间的桥梁，它对被测对象的拓扑结构依赖性不强，对被测系统的故障性质的敏感，可以单调地跟踪系统故障的变化，这种基于混沌特性分析的故障诊断方法与其它的模拟电路的故障诊断方法相比较具有显著的优势。

The conclusion that dissolved gases in gas composition and content of transformer could embody the inside of the transformer"s situation about "the insulation fault effectively is concluded.In the research of power equipment fault diagnosis based dissolved gases in transformer oil, the fault type increase clearly with the fault development. Traditional way is inability to faults diagnose of electrical equipment such as transformer that the fault reason, phenomenon and principle are indeterminacy etc.

在以油中溶解气体分析电气设备故障的模式识别研究中，当故障进一步发展，待判别的类型将显著增加，由于传统的诊断方法缺乏对样本数据的科学分析，不能够真实体现特征量数据的规律，将不可避免的面临准确率下降的问题，且不可能提供类间相互关系或演变过程的信息，因此用新的诊断方法来解决此类问题十分必要；本文在对目前变压器内部绝缘故障常用诊断方法总结分析的基础上，利用模糊数学的思想，提出以油中溶解气体为特征量，采用模糊聚类的方法对变压器内部绝缘故障进行诊断。

Cynanchum Lingtai apricot production in the average weight 65 grams, the brightly-colored fruit, juicy rich, sweet-sour taste, sweet from the nucleolus, when the late Qing Dynasty famous Shaanxi, Gansu provinces, the Qing imperial court Tongzhi tribute for years.

灵台生产的牛心杏平均单果重65克，果实色泽鲜艳，汁多味浓，甜酸适口，离核仁甜，清末时就驰名陕、甘两省，清同治年间曾为朝廷贡品。

Chenopodium album,Solanum nigrum, and Amaranthus retroflexus were very susceptible to the herbicides. Polygonum persicaria and Abutilon theophrasti were relatively less susceptible to the herbicides, and Lycopersicon esculentum was not susceptible to it. The relationship between reduction rates of weed biomass and PPM values of weed leaves 2,4, and 6 days after treatment was established.

供试的6种杂草对该混剂的敏感性存在显著差异：红心藜Chenopodium album、龙葵Solanum nigrum和反枝苋Amaranthus retroflexus对该混剂最敏感，ED90值分别为47.65、71.67和29.17g/hm2；春蓼Polygonum persicaria和苘麻Abutilon theophrasti敏感，ED90值分别为96.91、114.20g/hm2；而番茄不敏感。