量子ICA技术在故障诊断中的应用研究

发布时间：2018-01-18 04:20

本文关键词：量子ICA技术在故障诊断中的应用研究　出处：《中北大学》2012年硕士论文　论文类型：学位论文

【摘要】：近年来世界各国都对故障诊断十分重视，，因为一旦在生产过程中出现故障，都有可能使机械设备整体停工，给生产等带来极大的经济损失，甚至威胁工作人员的生命安全。所以，及时发现及排除故障具有很重要的现实意义。本文以齿轮箱为研究对象，将量子优化算法引入到盲源分离中，提出了量子独立分量分析算法（QICA算法），对齿轮箱产生的振动信号进行盲分离再提取其故障特征信息，最后用量子神经网络实现了对齿轮箱的故障诊断。本文主要工作内容如下：（1）分析并研究了独立分量分析算法和量子优化算法的理论、原理和重要公式的推导，详细研究了几种常用算法，为后续工作打基础。（2）创新性的提出了量子独立分量分析算法（QICA算法）。根据独立分量分析算法的一般求解过程，以量子算法为优化算法，提出了量子独立分量分析算法，并推导出算法中的关键理论，编制了算法程序。并通过仿真实验验证了该算法的可能性及其分离性能，结果表明量子优化应用于独立分量分析确实有助于提高算法的效率。（3）在对齿轮箱故障特征深入研究的基础上，利用齿轮箱在实验室设置的五种典型工况下测量和采集其振动信号，并使用量子独立分量分析算法（QICA算法）进行测试信号的盲源分离，提取其故障特征，再以量子神经网络进行齿轮箱的故障诊断。结果表明，经过量子盲源分离后，可提高齿轮箱的故障诊断的准确性和可靠性。
[Abstract]:In recent years, many countries all over the world attach great importance to the fault diagnosis, because once there is a fault in the production process, it is possible to make the whole machinery and equipment shut down, which brings great economic loss to the production and so on. Therefore, timely detection and troubleshooting of faults is of great practical significance. In this paper, a quantum optimization algorithm is introduced to blind source separation based on gearbox. A quantum independent component analysis (QICA) algorithm is proposed, in which the vibration signals generated by the gearbox are separated blind and the fault characteristic information is extracted. Finally, the fault diagnosis of gearbox is realized by quantum neural network. The main work of this paper is as follows: 1) the theory, principle and important formulas of independent component analysis (ICA) and quantum optimization algorithm are analyzed and studied, and several common algorithms are studied in detail to lay the foundation for further work. 2) an innovative quantum independent component analysis (QICA) algorithm is proposed. According to the general solution process of the independent component analysis (ICA) algorithm, the quantum algorithm is taken as the optimization algorithm. A quantum independent component analysis (QICA) algorithm is proposed, the key theory of the algorithm is deduced, and the algorithm program is developed. The possibility and separation performance of the algorithm are verified by simulation experiments. The results show that the application of quantum optimization to independent component analysis can improve the efficiency of the algorithm. On the basis of deeply studying the fault characteristics of the gearbox, the vibration signals of the gearbox are measured and collected under five typical working conditions set in the laboratory. The quantum independent component analysis (QICA) algorithm is used to separate the blind source of the test signal, extract its fault characteristics, and then use the quantum neural network to diagnose the gearbox fault. The accuracy and reliability of gearbox fault diagnosis can be improved by quantum blind source separation.
【学位授予单位】：中北大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH165.3

【参考文献】