基于EMD-WVD与LNMF的内燃机故障诊断
发布时间：2019-02-23 21:25
【摘要】：内燃机的振动信号是复杂非平稳信号,准确提取内燃机振动信号中的特征信息进行模式识别,是对振动信号进行故障诊断的关键。基于经验模态分解的维格纳时频分析方法,不但保留了维格纳分布的所有优良特,而且还避免了交叉项的干扰,能够有效地提取内燃机振动信号的特征信息;在此基础之上,针对传统非负矩阵分解非正交的基矩阵导致数据冗余性较大、影响后续故障分类准确率提高的问题,提出采用局部非负矩阵分解的方法,直接对EMD-WVD时频图像的矩阵进行分解,计算用于内燃机故障诊断的特征参数,并利用特征参数进行故障分类。对内燃机4种不同工况的振动信号进行实验,证明基于EMD-WVD与局部非负矩阵分解的方法对内燃机气门间隙的故障诊断的有效性。
[Abstract]:The vibration signal of internal combustion engine is a complex non-stationary signal. It is the key of fault diagnosis to accurately extract the characteristic information from the vibration signal of internal combustion engine for pattern recognition. The Wigner time-frequency analysis method based on empirical mode decomposition not only preserves all the excellent features of Wigner distribution, but also avoids the interference of crossover terms, and can effectively extract the characteristic information of internal combustion engine vibration signal. On this basis, a local nonnegative matrix decomposition method is proposed to solve the problem that the traditional non-negative matrix decomposition leads to greater data redundancy and affects the accuracy of subsequent fault classification. The matrix of EMD-WVD time-frequency image is decomposed directly, and the characteristic parameters used in internal combustion engine fault diagnosis are calculated, and the fault classification is carried out by using the characteristic parameters. The vibration signals of internal combustion engine under four different working conditions are tested and the results show that the method based on EMD-WVD and local nonnegative matrix decomposition is effective in the diagnosis of valve clearance of internal combustion engine.
【作者单位】： 第二炮兵工程大学五系;
【基金】：国家自然科学基金青年基金项目(51405498) 陕西省自然科学基金项目(2013JQ8023)
【分类号】：TK407
，
本文编号：2429217
[Abstract]:The vibration signal of internal combustion engine is a complex non-stationary signal. It is the key of fault diagnosis to accurately extract the characteristic information from the vibration signal of internal combustion engine for pattern recognition. The Wigner time-frequency analysis method based on empirical mode decomposition not only preserves all the excellent features of Wigner distribution, but also avoids the interference of crossover terms, and can effectively extract the characteristic information of internal combustion engine vibration signal. On this basis, a local nonnegative matrix decomposition method is proposed to solve the problem that the traditional non-negative matrix decomposition leads to greater data redundancy and affects the accuracy of subsequent fault classification. The matrix of EMD-WVD time-frequency image is decomposed directly, and the characteristic parameters used in internal combustion engine fault diagnosis are calculated, and the fault classification is carried out by using the characteristic parameters. The vibration signals of internal combustion engine under four different working conditions are tested and the results show that the method based on EMD-WVD and local nonnegative matrix decomposition is effective in the diagnosis of valve clearance of internal combustion engine.
【作者单位】： 第二炮兵工程大学五系;
【基金】：国家自然科学基金青年基金项目(51405498) 陕西省自然科学基金项目(2013JQ8023)
【分类号】：TK407
，
本文编号：2429217
本文链接：https://www.wllwen.com/kejilunwen/dongligc/2429217.html
