FAULT DIAGNOSIS OF ROTATING MACHINERY USING KNOWLEDGE-BASED FUZZY NEURAL NETWORK

A novel knowledge-based fuzzy neural network （KBFNN） for fault diagnosis is presented. Crude rules were extracted and the corresponding dependent factors and antecedent coverage factors were calculated firstly from the diagnostic sample based on rough sets theory. Then the number of rules was used to construct partially the structure of a fuzzy neural network and those factors were implemented as initial weights, with fuzzy output parameters being optimized by genetic algorithm. Such fuzzy neural network was called KBFNN. This KBFNN was utilized to identify typical faults of rotating machinery. Diagnostic results show that it has those merits of shorter training time and higher right diagnostic level compared to general fuzzy neural networks.

A Fault Diagnosis Expert System for a Heavy Motor Used in a Rolling Mill

A fault diagnosis expert system for a heavy motor used in a rolling mill is established in this paper. The fault diagnosis knowledge base was built, and its knowledge was represented by production rules. The knowledge base includes daily inspection system, brief diagnosis system and precise diagnosis system. A pull down menu was adopted for the management of the knowledge base. The system can run under the help of expert system development tools. Practical examples show that the expert system can diagnose faults rapidly and precisely.

Synthetic Intelligent Fault Diagnosis Technology for Complex Process

A fault diagnosis method of knowledge based fuzzy neural network is proposed for complex process, which is hard to develop practical mathematical model. Fault detection is performed through a knowledge based system, where fault detection heuristic rules have been generated from deep and shallow knowledge of the process. The fuzzy neural network performs the fault diagnosis task. This method does not need practical mathematical models of objects, so it is a strong implement for complex process.

Continual learning fault diagnosis:A dual-branch adaptive aggregation residual network for fault diagnosis with machine increments

As a data-driven approach, Deep Learning(DL)-based fault diagnosis methods need to collect the relatively comprehensive data on machine fault types to achieve satisfactory performance. A mechanical system may include multiple submachines in the real-world. During condition monitoring of a mechanical system, fault data are distributed in a continuous flow of constantly generated information and new faults will inevitably occur in unconsidered submachines, which are also called machine increments. Therefore, adequately collecting fault data in advance is difficult. Limited by the characteristics of DL, training existing models directly with new fault data of new submachines leads to catastrophic forgetting of old tasks, while the cost of collecting all known data to retrain the models is excessively high. DL-based fault diagnosis methods cannot learn continually and adaptively in dynamic environments. A new Continual Learning Fault Diagnosis method(CLFD) is proposed in this paper to solve a series of fault diagnosis tasks with machine increments. The stability–plasticity dilemma is an intrinsic issue in continual learning. The core of CLFD is the proposed Dual-branch Adaptive Aggregation Residual Network(DAARN).Two types of residual blocks are created in each block layer of DAARN: steady and dynamic blocks. The stability–plasticity dilemma is solved by assigning them with adaptive aggregation weights to balance stability and plasticity, and a bi-level optimization program is used to optimize adaptive aggregation weights and model parameters. In addition, a feature-level knowledge distillation loss function is proposed to further overcome catastrophic forgetting. CLFD is then applied to the fault diagnosis case with machine increments. Results demonstrate that CLFD outperforms other continual learning methods and has satisfactory robustness.

A new interpretable fault diagnosis method based on belief rule base and probability table

It is vital to establish an interpretable fault diagnosis model for critical equipment.Belief Rule Base(BRB)is an interpretable expert system gradually applied in fault diagnosis.However,the expert knowledge cannot be utilized to establish the initial BRB accurately if there are multiple referential grades in different fault features.In addition,the interpretability of BRB-based fault diagnosis is destroyed in the optimization process,which reflects in two aspects:deviation from the initial expert judgment and over-optimization of parameters.To solve these problems,a new interpretable fault diagnosis model based on BRB and probability table,called the BRB-P,is proposed in this paper.Compared with the traditional BRB,the BRB-P constructed by the probability table is more accurate.Then,the interpretability constraints,i.e.,the credibility of expert knowledge,the penalty factor and the rule-activation factor,are inserted into the projection covariance matrix adaption evolution strategy to maintain the interpretability of BRB-P.A case study of the aerospace relay is conducted to verify the effectiveness of the proposed method.

Knowledge Processing Method of Fault Diagnosis Expert Systems for Letter Sorting Equipment

Based on the analysis of fault diagnosis knowledge of letter sorting machine, this paper proposes a processing method by which the fault diagnosis knowledge is divided into exact knowledge, inadequate knowledge and fuzzy knowledge. Then their presenting and implementing form in fault diagnosis expert system is discussed and studied. It is proved that the expert system has good feasibility in the field of the diagnosis of letter sorting machine.

Study on Fault Diagnosis of Rotating Machinery with Hybrid Neural Networks

With the help of the feedforward neural network diagnostic method, the hybrid diagnostic networks corresponding to information in multiple symptom domains are built and the comprehensive judgment is carried out with weighted average method. Meanwhile, this method has the ability of self learning and self adaptation in order to adapt both the complexity of vibrations produced practically and the pluralistic potent of vibration symptoms induced really for large rotating machinery, especially for turbogenerators. The reliability and precision of diagnosis with this method is heightened. It seems that the method can take more practical value in engineering applications.

基于Adaboost-INGO-HKELM的变压器故障辨识

针对目前变压器故障诊断准确率低的问题,提出一种多策略集成模型。首先通过等度量映射(isometric mapping, Isomap)对高维非线性不可分的变压器故障数据进行降维处理。其次,利用混合核极限学习机(hybrid kernel based extreme learning machine, HKELM)进行训练学习,考虑到HKELM模型易受参数影响,所以利用北方苍鹰优化算法(northern goshawk optimization, NGO)对其参数进行寻优。但由于NGO收敛速度较慢,易陷入局部最优,引入切比雪夫混沌映射、择优学习、自适应t分布联合策略对其进行改进。同时为了提高模型整体的准确率,通过结合Adaboost集成算法,构建Adaboost-INGO-HKELM变压器故障辨识模型。最后,将提出的Adaboost-INGO-HKELM模型与未进行降维处理的INGO-HKELM模型、Isomap-INGO-KELM模型、Adaboost-Isomap-GWO-SVM等7种模型的测试准确率进行对比。提出的Adaboost-INGO-HKELM模型的准确率可达96%,均高于其他模型,验证了该模型对变压器故障辨识具有很好的效果。

基于电子知识库的船舶故障诊断技术及应用

针对船舶电子设备故障的诊断问题,提出一种基于知识图谱的故障诊断系统。该系统不仅利用传统的故障维修日志库,还利用图书馆电子数据源提取和整合广泛的知识,建立起一个详尽的船舶设备故障知识图谱,为运维人员提供了一个全面的参考框架,并使用机器学习方法实现对电子设备故障的快速和准确诊断,显著提高了船舶故障处理的效率。

多策略改进蜉蝣算法在变压器故障诊断中的应用

为提高支持向量机在变压器故障诊断的准确率,提出了一种多策略改进蜉蝣算法优化支持向量机的故障诊断方法,并通过利用螺旋函数、正余弦自适应权重优化改进后的蜉蝣算法,得到支持向量的最佳参数c和g。通过3个测试函数对改进后的算法进行仿真对比,实验表明改进后的算法具有较高的寻优精度和较快的收敛速度。将提出的故障诊断方法运用到实际DGA故障数据中,结果表明该方法能有效提高变压器故障诊断的准确率和缩短运行时间。

基于EBWO-SVM的变压器故障诊断研究

针对SVM在变压器故障诊断中存在诊断精度不高和BWO算法存在易陷入局部最优的问题,提出一种基于EBWO-SVM的变压器故障诊断方法。首先通过引入准反向学习策略和旋风式觅食策略对BWO算法进行改进,然后将EBWO算法与粒子群优化算法、灰狼优化算法、鲸鱼优化算法、白鲸优化算法在6种测试函数上进行寻优测试,验证了EBWO算法的优越性。其次利用EBWO算法对SVM中的核函数参数g和C进行优化,从而提高SVM的分类能力。最后提出其他方法与EBWO-SVM模型进行对比。结果表示:所构建的EBWO-SVM变压器故障诊断模型与BWO-SVM、WOA-SVM、GWO-SVM、PSO-SVM相比,综合诊断精度分别提高了7.7%、9.7%、11.6%、15.4%,且稳定性更强,验证了EBWO-SVM模型的可行性与有效性。

基于RegNet-CSAM与ZOA-KELM模型的滚动轴承故障诊断

针对现有深度卷积神经网络对滚动轴承混合故障诊断效果不佳以及模型复杂度过高导致计算成本过大等问题,提出了一种基于RegNet-CSAM与ZOA-KELM模型的滚动轴承故障诊断方法。该模型由RegNet-CSAM网络和ZOA-KELM分类算法组成。首先,将融合了通道和空间特征的注意力机制CSAM与组卷积残差模块结合,提升该结构的表征能力,由此构建的RegNet-CSAM网络,模型复杂度为0.48GF;其次,在分类阶段将斑马优化核极限学习机(ZOA-KELM)替代原来网络中使用的Softmax函数完成最后的分类任务。滚动轴承故障诊断试验结果表明,RegNet网络对滚动轴承混合故障样本容易产生误判,CSAM的融入虽将RegNet网络的分类精度进一步提高,但是仍然存在一定程度的滚动轴承混合故障误判问题;而将ZOA-KELM替代Softmax函数后再对RegNet-CSAM网络输出特征进行分类,能够有效识别出滚动轴承的单一和混合故障,准确率达到了99.92%。所提方法对比其他网络,诊断精度最大提升5.02%,模型复杂度最大缩减32倍。

基于对抗神经网络的力学试验机故障诊断系统设计

针对高校力学试验机随机故障频发问题,设计了力学试验机故障诊断系统。针对此类试验机使用间歇长、离散性强等特征,提出了基于离散性数据识别故障的设计思路。建立了机器学习知识库,基于对抗神经网络(GAN)理论设计了力学试验机故障诊断算法并建立了故障诊断系统,评价了故障诊断系统的性能指标。结果表明,所建立故障诊断系统在实验中的最低精确率、准确率和召回率分别达到96.12%、96.51%和95.15%,最高误识率仅为3.96%,性能满足使用要求。

基于TrAdaBoost的冷水机组故障迁移诊断

冷水机组故障可以通过机器学习进行诊断,但需大量训练数据,而获取有效故障数据难度大、成本高。传统故障诊断主要针对单台机组已有数据,很难覆盖全部工况,新工况下诊断性能恶化。本文提出了用对数据进行空间挤压的多重数据处理方法缩减不同分布间的差异,并利用TrAdaBoost算法对不同数据分布的信息迁移能力,结合不同基分类器搭建了冷水机组故障诊断模型,实现了新工况故障的有效诊断,有望缩减实验成本。对冷水机组7类典型故障的诊断结果显示:在新工况数据仅为20组时,相比于未进行迁移诊断的情况,总体正确率分别提升了22.00%、2.50%和32.33%。通过增补2个工况数据验证了不同模式下迁移诊断对冷水机组故障诊断的有效性:单模式下迁移诊断性能较常规诊断提高18.39%~22.43%,全模式下提高1.21%~2.55%;参数寻优对单模式迁移诊断有辅助提升效果(3.06%),对全模式则因过拟合导致性能下降(-4.23%)。可见,基于源工况知识与目标工况少量数据的迁移诊断模型是解决新工况数据缺乏问题的有效途径。

基于机器阅读理解的行车故障诊断知识抽取

行车故障调查单是对行车故障诊断过程的文本记录,基于这些历史记录构建知识图谱可以更好地支持行车故障诊断智能化。由于该语料具有实体嵌套、实体跨度大、关系重叠等特点,传统的命名实体识别和关系抽取模型难以对其进行有效的知识抽取。针对语料中存在的实体嵌套和长实体识别问题,本文提出了一种融合强化学习的机器阅读理解模型,以问答形式进行实体识别,以指针网络进行解码;对于语料中存在的关系重叠问题,将关系抽取分为先识别主体再识别客体的两阶段,将不同实体对的关系抽取进行隔离。实验结果表明,基于机器阅读理解的方法在行车故障诊断领域的知识抽取上具有较好的性能,可以有效支持领域知识图谱构建。

城市集中供热水处理专家帮助系统的开发

随着智能城市的不断发展,专家帮助系统在人们的生活中发挥着至关重要的作用。针对热电联产机组和热水管网水处理经常出现的一些故障进行分析,将热源厂水处理中常出现的设备故障问题及其解决方案制作成数据库和专家知识库,在windows环境下,将数据库与Visual Basic6.0软件对接,开发出供水管网故障诊断专家帮助系统软件。主要功能有查询数据库中的供热管网故障类型、解决方案、即时更新数据库内容。

基于数据驱动和本体建模的数控机床主轴故障诊断与推理

针对目前数控机床主轴系统故障诊断存在方法单一及智能化程度低的问题,提出基于数据驱动和本体建模的机床主轴故障诊断与推理方法。采用EMD对传感器采集的蕴含故障特征的原始信号进行数据处理与分析,提取原始统计特征,在此基础上,构建DBN-RF诊断模型实现深度特征自适应挖掘与故障模式识别。利用Protégé5.1工具结合领域知识构建机床主轴故障本体知识库,将DBN-RF诊断模型的故障辨识结果与本体知识库中的实例进行语义映射,实现故障知识推理,获得故障原因和故障解决策略。基于采集的不同工况下轴承故障数据验证了DBN-RF诊断模型的有效性,最高故障诊断平均准确率可达92.93%;构建实例验证了本体知识库的可重用性和推理功能;最后,设计开发了数控机床主轴健康管理服务系统,实现主轴系统状态实时感知和故障诊断与推理。

基于卫星宽带的远程设备监测诊断技术研究

针对船(岛)上设备信息系统存在故障诊断不够智能化以及系统可扩展性不足等问题,提出一种利用卫星通道的远程监测诊断技术并实现其系统,该系统基于物模型规范化管理异构设备信息,船(岛)侧基于设备网关管理所有数据出入,支持利用知识库与测试集进行故障分析推理,能够很好地支撑岸端技术人员对船上设备进行智能诊断;物联网设备数据统一接入模式有效提升了系统的可扩展性。

UNCERTAIN KNOWLEDGE MANAGEMENT IN EXPERT SYSTEMS USING FUZZY METAGRAPHS

This paper presented a new graph theoretic construct——fuzzy metagraphs and discussed their applications in constructing fuzzy knowledge base. Fuzzy metagraphs describe the relationships between sets of fuzzy elements but not single fuzzy element and offer some distinct advantages both for visualization of systems, as well as for formal analysis of system structure. In rule based system, a fuzzy metagraph is a unity of the knowledge base and the reasoning engine. Based on the closure of the adjacency matrix of fuzzy metagraphs, this paper presented an optimized inferential mechanism working mainly by an off line approach. It can greatly increase the efficiency of inference. Finally, it was applied in a daignostic expert system and satisfactory results were obtained.

基于改进层次基本熵融合SMA-SVM模型的轴承故障诊断方法

针对煤矿机械轴承的故障特征提取和故障状态识别问题,提出了改进层次基本熵(IHBSE)特征提取融合黏菌优化(SMA)—支持向量机(SVM)分类模型的煤矿机械轴承故障诊断方法。首先,引入了能够同时分析信号低频和高频信息的IHBSE方法,并将其用于捕捉不同状态下,煤矿机械轴承振动信号中的多维故障特征,构建了特征向量;然后,采用具有优异全局寻优性能的黏菌算法,对支持向量机的惩罚系数和核函数的最佳值进行了搜索,提出了黏菌算法—支持向量机(SMA-SVM)模型;最后,利用部分特征样本对诊断模型进行了训练,并采用训练完毕的具有最佳参数的SMA-SVM分类器,进行了轴承故障类型和严重程度的判断。研究结果表明:所提出的煤矿机械轴承故障诊断方法可以有效地识别煤矿机械轴承的运行状态,分类准确率达到了1,而在多次实验下的平均准确率也高于0.98,对实际工程应用具有一定的参考价值。