共查询到20条相似文献,搜索用时 62 毫秒
1.
小波包特征熵神经网络在尾水管故障诊断中的应用 总被引:26,自引:5,他引:26
为精确诊断水轮机尾水管涡带,该文提出一种基于小波包特征熵的神经网络故障诊断新方法。对采集到的尾水管压力脉动信号进行三层小波包分解,提取小波包特征熵,然后构造信号的小波包特征熵向量,并以此向量作为故障样本对三层BP神经网络进行训练,实现智能化故障诊断。试验结果表明训练成功的BP网络能够很好地诊断机组尾水管是否发生涡带以及涡带的严重程度,为水轮机故障诊断开辟新的途径。 相似文献
2.
3.
小波包与神经网络在电机故障诊断中的应用研究 总被引:1,自引:0,他引:1
文章对电机的故障特点进行分析,根据小波包变换能将信号按任意时频分辨率分解到不同频段的特性,结合小波包的能量特性,提出了故障信号在不同分解频段的能量特征概念及算法,并将其与BP神经网络相结合,提出了一种新的电机故障诊断方法,实验结果证实了该方法的正确性和有效性。 相似文献
4.
轴承是旋转机械设备的关键部件,目前已有很多轴承故障诊断方法,但其中一些方法只能针对特定的轴承故障进行诊断,可能不适用于其他轴承故障问题,而且大部分方法的诊断准确率还可以进一步提高。提出小波包能量熵与深度置信网络(DBN)相结合的方法进行轴承故障诊断。首先对轴承振动信号进行小波包变换,然后以能量熵的形式构建特征向量,这些特征向量含有不同频段内的振动能量大小,可以用于区分各种轴承故障。最后利用基于DBN的深度模型对能量熵特征向量进行故障识别。使用两类轴承数据集进行验证,分别获得100%和99.5%的故障识别准确率。实验结果表明,该诊断方法具有较好的通用性,而且可以达到很高的诊断准确率。 相似文献
5.
小波熵证据的信息融合在电力系统故障诊断中的应用 总被引:9,自引:1,他引:8
电力系统中快速准确的故障诊断是事故后隔离故障元件、恢复系统正常运行的首要前提,具有重要意义。该文从信息融合的角度出发,提出利用多种小波熵测度的融合来解决电力系统故障诊断问题。小波熵测度由于结合了小波变换和信息熵理论的优势,能快速准确地提取线路故障特征,但由于故障的不确定性和多样性,依靠单一的小波熵测度诊断故障可能出现诊断困难或诊断失真等问题,因此提出采用D-S证据理论对多种小波熵进行信息融合,并采用范数加权平均的方法来建立基本可信度分配,以基本可信数的决策方法来实现故障模式诊断。基于EMTDC和Matlab的仿真证明,该方法能提高对故障诊断结果的支持度及故障诊断的准确性和实时性,是故障模式定量诊断的一种可行性新方法。 相似文献
6.
7.
针对轴承故障诊断时振动信号呈现复杂性和混沌特性,故障特征分量容易淹没在噪声之中。引用自适应线性神经网络(Adaptive Linear Neuron,ADALINE)降噪和小波包Shannon熵(Wavelet Packet Analysis Shannon Entropy,WPASE)相结合的方法诊断轴承故障。首先利用ADALINE对不同故障模式的振动信号进行降噪处理,引用小波包理论对降噪后的信号进行小波包分解,计算各层细节信号的Shannon熵值,以此作为不同故障模式的故障特征量。仿真实验表明ADALINE降噪效果明显,Shannon熵能够清楚区别不同的故障模式。该方法简单可靠,为轴承故障诊断提供了新的思路和方法。 相似文献
8.
9.
10.
11.
12.
13.
S. El Safty A. El-Zonkoly 《International Journal of Electrical Power & Energy Systems》2009,31(10):604-607
The ability to detect and classify the type of fault plays a great role in the protection of power system. This procedure is required to be precise with no time consumption. In this paper detection of fault type has been implemented using wavelet analysis together with wavelet entropy principle. The simulation of power system is carried out using PSCAD/EMTDC. Different types of faults were studied obtaining various current waveforms. These current waveforms were decomposed using wavelet analysis into different approximation and details. The wavelet entropies of such decompositions are analyzed reaching a successful methodology for fault classification. The suggested approach is tested using different fault types and proven successful identification for the type of fault. 相似文献
14.
15.
联合AGR的神经网络在电力系统故障和振荡识别中的应用 总被引:2,自引:0,他引:2
结合最优联合时一频处理无交叉项干扰及神经网络自学习分类识别的优点,提出了一种在有色噪声干扰下识别电力系统故障和振荡的方法。将经过自适应高斯基表示(Adaptive Gaussian Representation,AGR)分析处理的电力信号特征向量输入神经网络分类器进行识别。待辨识输入向量不仅表征了原信号的基本信息,而且没有交叉项,运算简单。仿真结果表明,此方法能正确分类识别有色噪声干扰下的系统故障和振荡,提高了电力系统微机保护在系统振荡中检测故障的灵敏性和精确性。 相似文献
16.
17.
18.
19.
This paper presents a novel wavelet based approach for fault location using voltage transient waveforms in power distribution systems. The proposed method includes two main stages. Firstly, the approximate location of the fault or fault section is determined using a new algorithm with discrete wavelet transform. The difference between arriving times of transient components in different measurement units is used for this purpose. The accurate location of the fault is determined in the second stage. Depending on the determined fault section, the difference between arriving times of transient components in different measurement units or the frequency content of the voltage transients are used. The time difference and frequency content are calculated using discrete and continuous wavelet transform (DWT and CWT) respectively. The proposed technique is implemented on an unbalanced 34 bus distribution system with two distributed generation units which is simulated in ATP–EMTP. The comparison of the results of the proposed method with previous works verifies its better accuracy and more robustness to fault conditions including fault inception angle and fault resistance. 相似文献
20.
This paper proposes a new method of fault detection and classification in asymmetrical distribution systems with dispersed generation to detect islanding and perform protective action based on applying a combination of wavelet singular entropy and fuzzy logic. In this method, positive components of currents at common coupling points are decomposed to adjust detailed coefficients of wavelet transforms and singular value matrices, and expected entropy values are calculated via stochastic process. Indexes are defined based on the wavelet singular entropy in positive components and three phase currents to detect and classify the fault. This protection scheme is put forward for fault detection and is investigated in different types of faults such as single-phase to ground, double-phase to ground, three-phase to ground and line to line in distribution lines in the presence of distributed generations, and different locations of faults are verified when the distributed generation is connected to the utility. The major priority of the proposed protection scheme is its reduction in time (10 ms from the event inception) in distinguishing islanding and protection transmission lines in the presence of distributed generations. 相似文献