| 集成相空间重构与稀疏自编码的振动信号分解方法 |
| |
| 引用本文： | 杜灿谊,林祖胜,喻菲菲,张绍辉.集成相空间重构与稀疏自编码的振动信号分解方法[J].机械设计与制造,2019(8):67-72. |
| |
| 作者姓名： | 杜灿谊 林祖胜 喻菲菲 张绍辉 |
| |
| 作者单位： | 广东技术师范大学汽车与交通工程学院,广东 广州,510665;厦门理工学院机械与汽车工程学院,福建 厦门,361024;东莞理工学院机械工程学院,广东 东莞,523808 |
| |
| 基金项目： | 国家自然科学基金;福建省中青年教师教育科研项目;广东省自然科学基金 |
| |
| 摘 要： | 稀疏自编码算法通过对输入信号的编码与解码过程使得输出信号能够最大程度的保留输入数据信息,具备强大的数据处理功能,然而,SAE的输出数据具有什么特性、噪声成分在转化过程中发生怎样的变化并没有详细研究,针对该问题,以时域振动信号作为输入,理论推导噪声在编码与解码过程中的变化,分析输出信号的具体成分,得出信号在转化过程中能够有效的滤除噪声成分,并分离出信号的主要成分。仿真振动数据及齿轮箱故障诊断实验证明,稀疏自编码算法能够有效的提高振动信号的信噪比和提取故障的频率成分。
|
| 关 键 词： | 稀疏自编码 信号处理 降噪 |
Vibration Signal Decomposition Based on Phase Space Reconstruction and Sparse Auto-Encoding |
| |
| DU Can-yi,LIN Zu-sheng,YU Fei-fei,ZHANG Shao-hui.Vibration Signal Decomposition Based on Phase Space Reconstruction and Sparse Auto-Encoding[J].Machinery Design & Manufacture,2019(8):67-72. |
| |
| Authors: | DU Can-yi LIN Zu-sheng YU Fei-fei ZHANG Shao-hui |
| |
| Affiliation: | (School of Automotive and Transportation Engineering, Guangdong Polytechnic Normal University, Guangdong Guangzhou 510665, China;School of Mechanical and Automotive Engineering,Xiamen University of Technology, Fujian Xiamen 361024, China;School of Mechanical Engineering, Dongguan University of Technology, Guangdong Dongguan 523808,China) |
| |
| Abstract: | The sparse auto-encoding algorithm can preserve the information of the input data by maximizing the output data through the encoding and decoding process of the input data,and has a powerful data processing function.However,what is the characteristic of the output data of SAE and what changes in the transformation process of the noise components have not been studied in detail.In this paper,the time domain vibration signal is used as input,the change of noise in the encoding and decoding process is derived,and the specific components of the output signal are analyzed.The noise components can be filtered effectively during the conversion process,and the main components of the signal are separated.The simulation vibration data and the gear box fault diagnosis experiment prove that the sparse auto-encoding algorithm can effectively improve the signal to noise ratio of the vibration signal and the frequency component of the fault extraction. |
| |
| Keywords: | Sparse Auto-Encoding Signal Processing Denoising |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
