共查询到17条相似文献,搜索用时 171 毫秒
1.
针对旋转机械故障率偏高,而人工参与故障诊断工作量大、效率偏低等问题,提出一种基于云模型与LSTM算法的旋转机械故障诊断方法。采用实验台采集振动故障原始数据,统一进行EEMD数据预处理,利用云模型进行故障特征数据提取,输入LSTM神经网络模型进行故障诊断。通过云模型和能量法进行特征提取,分别输入支持向量机和LSTM神经网络模型进行诊断结果对比。结果表明:云模型与LSTM算法的故障诊断准确率最高,达到98.75%,证明该方法能够有效应用在旋转机械故障诊断中。 相似文献
2.
3.
时域中的量纲一指标因对故障敏感,被广泛运用于机械故障诊断中,但是目前量纲一指标在诊断过程中存在严重交叉问题,即量纲一指标对于不同故障状态在特征空间中存在混叠现象。为了解决这个问题,提出基于量纲一指标和极限学习机的滚动轴承故障识别方法,采用美国西储大学轴承数据中心网站公开发布的轴承探伤数据集,验证算法诊断效果。为了进一步验证算法的优越性,将该算法与BP神经网络、支持向量机(SVM)和Grip search SVM 3种算法进行比较,结果表明:基于量纲一指标和极限学习机的故障诊断方法能够提高滚动轴承故障诊断效率和分类准确率。 相似文献
4.
5.
基于机器学习故障诊断方法,针对船用滚动轴承复合故障特征提取多样化的特点,提出一种以振动信号时域指标为特征的随机森林故障诊断方法。将振动时域信号进行清洗转换,构造5个量纲一化指标的衍生特征,并选取以决策树为基本分类器的随机森林算法建立训练模型;通过特征筛选、评估测试和模型优化得到较为理想的故障诊断分类模型;采用滚动轴承竞赛数据集进行模型仿真,并结合实际模拟8种船用滚动轴承故障状态。通过三向振动实验和算法建模,证明特征提取的科学性和故障诊断模型的有效性。结果表明:采用该方法,数据仿真诊断准确率为98.61%,实验诊断准确率为98.85%,且该方法在振动采集方向为轴向时诊断效果最优。 相似文献
6.
多传感器信息融合在液压系统智能故障诊断中的应用 总被引:2,自引:0,他引:2
针对采矿工程机械液压系统故障诊断方法存在的局限性,提出了一种基于多传感器信息融合的智能故障诊断方尊。该方法采用模糊神经网络融合诊断中心作为故障诊断的执行机构,算法上采用BP算法。通过一实例论证了在液压系统故障诊断中采用多传感器信息融合故障诊断方法比采用单传感器信息故障诊断方法更具有准确性和可靠性。 相似文献
7.
8.
9.
针对旋转机械轴承微弱故障振动信号易被强噪声掩盖难以识别的问题,提出一种改进混沌粒子群优化支持向量机的故障诊断方法。将信号通过局部均值分解算法分解处理得到乘积函数(PF)分量,并进行能量归一化处理获得时频域特征集;通过迭代拉普拉斯得分降低时频域特征集的空间维度;以PF分量的排列熵作为混沌粒子群的适应度,并加入交叉和变异新策略,建立一种新的交叉变异混沌粒子群优化方法;利用改进的粒子群算法优化支持向量机的核函数和惩罚因子,并将优化后的分类模型应用于轴承故障诊断。结果表明:该故障分类模型的识别准确率高于其他分类模型。 相似文献
10.
基于CBR的旋转机械故障诊断专家系统的设计 总被引:1,自引:0,他引:1
针对旋转机械故障诊断技术的特点,文章在综合比较多种知识推理技术的基础上,决定采用基于分阶段近邻检索算法的基于案例推理技术(CBR),同时运用Access数据库构建案例库存储结构,并结合VC++编程环境,构建了旋转机械故障诊断专家系统.最后通过旋转机械故障诊断实验台,对专家系统诊断结果进行了验证.实验结果表明,该系统具有诊断准确度高,诊断速度快,易维护等特点,能够充分利用已有的案例知识,有效地协助技术人员解决设备故障,提高诊断效率. 相似文献
11.
传统故障诊断方法依赖于先验数据与模型,具有局限性。为解决此问题,提出一种基于数据驱动的旋转机械故障诊断方法。利用经验模式分解(EMD)算法拆分原始故障信号,得到有限个IMF分量,优化现有EMD算法得到最优的截断阈值,并有效分离系统噪声干扰;从多域量化角度提取故障信号的时域、频域特征,并基于EMD样本熵实现对去噪旋转机械故障信号中故障点特征的分类与识别。仿真结果表明:所提出的数据驱动算法能够准确地识别出不同载荷条件下的故障信号微弱特征,具有更高的训练精度和故障诊断精度 相似文献
12.
针对旋转机械监测中无法随时随地查看其运行状态,监测产生的数据量逐渐加大,以及故障特征提取困难的问题,以轴承作为关键部件,提出一种基于云平台的旋转机械轴承监测系统。系统采用温度和加速度传感器、STM32单片机获得轴承监测所需的数据;然后利用窄带物联网完成数据远程传输,并将其存储到云端数据库中;在云平台利用相关时域频域分析对轴承状态进行监测,并利用设计的一种多尺度一维卷积神经网络模型实现轴承的故障诊断;然后由Web浏览器显示轴承的运行状态和故障诊断结果。实验结果表明提出的故障诊断方法诊断准确率高、效果好,系统能够良好地运行。 相似文献
13.
在机械故障识别方面,因子隐Markov模型是目前常用的识别工具。无限因子隐Markov模型(IFHMM)是因子隐Markov模型(FHMM)的一种扩展形式,克服了因子隐Markov模型链条数往往事先假定的缺点。本研究将无限因子隐Markov模型(IFHMM)运用到旋转机械的升降速过程故障的诊断当中,提出了使用IFHMM作为诊断工具的旋转机械故障诊断方法,并与基于因子隐Markov模型的旋转机械故障诊断方法进行了对比,最后将提出的方法成功地应用到旋转机械的故障中。实验结果表明,提出的方法明显优于FHMM识别方法。 相似文献
14.
An order-tracking technique for the diagnosis of faults in rotating machineries using a variable step-size affine projection algorithm 总被引:3,自引:0,他引:3
This paper describes principles and applications of an adaptive order tracking diagnosis technique for fault diagnosis in rotating machinery. An adaptive high-resolution order tracking method with a variable step-size affine-projection algorithm (VSS APA) is used to diagnose faults in the gear-set and centrifugal fan blades. In comparison with conventional order-tracking methods such as recursive least-square filtering algorithm, the proposed VSS APA technique has fast convergence speed for adaptive filtering process. The VSS APA-based order-tracking technique can overcome problems encountered in FFT based methods. The smearing problem is treated as the tracking of frequency-varying band-pass signals. Ordered amplitudes can be calculated with high-resolution adaptive filter algorithm after experimental implementations carried out to evaluate the proposed algorithm in gear-set and centrifugal fan blades defect diagnosis. The experimental result indicates that the proposed algorithm is effective in rotating machinery fault diagnosis. 相似文献
15.
Design of an adaptive line enhancement system using a variable step-size affine-projection algorithm
This paper presents design of an adaptive line enhancement (ALE) system for improving sensor response using a variable step-size affine-projection algorithm (VSS APA). ALE is an adaptive technique that may be used to detect a periodic signal buried in a broadband noise background such as in rotating machinery fault diagnosis. However, most of the conventional methods for ALE system are based primarily on an adaptive filter with the least-mean-square (LMS) error algorithm. Unfortunately, convergence speed is limited when a filtering plant is varied, because the learning process of the adaptive algorithm fails to respond quickly enough to the changing operational conditions. This study proposed a VSS APA for improving both the convergence speed and the performance of the ALE system. Two applications were conducted to compare the performance of the proposed algorithm and various traditional adaptive filtering algorithms. The first application used the proposed ALE system to improve the response of a wheel speed sensor output signal; the other was used for reducing the background noise during rotating machinery fault diagnosis. Both the experimental results indicated that the ALE with VSS APA has an effective performance and convergence for both applications. 相似文献
16.
An application of a recursive Kalman filtering algorithm in rotating machinery fault diagnosis 总被引:3,自引:1,他引:3
In this paper, an application of adaptive order tracking fault diagnosis technique based on recursive Kalman filtering algorithm is presented. Order tracking fault diagnosis technique is one of the important tools for fault diagnosis of rotating machinery. Conventional methods of order tracking are primarily based on Fourier analysis with reference to shaft speed. In this study, a high-resolution order tracking method with adaptive Kalman filter is used to diagnose the fault in a gear set and damaged engine turbocharger wheel blades. The adaptive Kalman filtering algorithm can overcome the problems encountered in conventional methods. The problem is treated as the tracking of frequency-varying bandpass signals. Ordered amplitudes can be calculated with high resolution after experimental implementation. Experiments are also carried out to evaluate the proposed system in gear-set defect diagnosis and engine turbocharger wheel blades damaged under various conditions. The experimental results indicate that the proposed algorithm is effective in fault diagnosis of both cases. 相似文献
17.