基于自适应波形分解与时频转换的风机工况诊断方法
黄子恒1，许钊源1，方辉1，伍剑波*,1，李晋航2，石致远2
（1.四川大学 机械工程学院，四川 成都 610065；2.东方电气集团 中央研究院，四川 成都 611731）
摘要：目前针对风力发电机工况诊断方法的研究较为匮乏。为此，本文提出一种基于自适应波形分解和时频转换的诊断方法。该方法的算法采用了滑动间隔样本扩充、基于改进CEEMDAN的自适应波形分解方法、希尔伯特时频变换、VGG16神经网络。过程中，利用风机的振动加速度信号，通过自适应分解方法对信号进行分解降噪，将处理后信号进行变换生成二维时频谱，利用VGG16模型对时频谱样本集进行分类以达到诊断目的。为了评估该方法的有效性，将本文方法分别与传统模态分解、CNN模型进行对比，并对模型分类效果进行评估，结果表明，该方法具有更好的准确性。
关键词：风力发电机；波形分解；时频谱；VGG16
中图分类号：TK83 文献标志码：A doi：10.3969/j.issn.1006-0316.2023.10.004
文章编号：1006-0316 (2023) 10-0020-08
A Method for Wind Turbine Condition Diagnosis Based on Adaptive Decomposition and Time-Frequency Transform
HUANG Ziheng1，XU Zhaoyuan1，FANG Hui1，WU Jianbo1，LI Jinhang2，SHI Zhiyuan2
( 1.School of Mechanical Engineering, Sichuan University, Chengdu 610065, China; 2.The Research and Development Center, Dongfang Electric Corporation, Chengdu 611731, China )
Abstract：At present, the research on the working condition diagnosis method of wind turbines is relatively lacking. In this paper, a diagnosis method based on adaptive decomposition and time-frequency transform is proposed. The algorithm of this method adopts the sliding interval sample expansion, the adaptive waveform decomposition method based on improved CEEMDAN, the Hilbert transform, and the VGG16 neural network. During the process, firstly, the vibration acceleration signal is decomposed and denoised by the adaptive decomposition method, the processed signal is transformed to generate a two-dimensional time spectrum, and the VGG16 model is used to classify the time spectrum set. In order to evaluate the effectiveness of the method, the method in this paper is compared with the traditional decomposition mode and CNN model. The performance of the classification is evaluated and the result shows that the method has better accuracy.
Key words：wind turbine；waveform decomposition；time spectrum；VGG16
———————————————
收稿日期：2022-09-23
基金项目：四川省科技计划重点研发项目（2021YFG0039）
作者简介：黄子恒（1996－），男，四川成都人，硕士研究生，主要研究方向为智能控制与故障诊断，E-mail：1738568615@qq.com。*通讯作者：伍剑波（1986－），男，四川眉山人，博士，教授、博士研究生导师，主要研究方向为智能控制与无损检测，E-mail：wujianbo@scu.edu.cn。
（1.四川大学 机械工程学院，四川 成都 610065；2.东方电气集团 中央研究院，四川 成都 611731）
摘要：目前针对风力发电机工况诊断方法的研究较为匮乏。为此，本文提出一种基于自适应波形分解和时频转换的诊断方法。该方法的算法采用了滑动间隔样本扩充、基于改进CEEMDAN的自适应波形分解方法、希尔伯特时频变换、VGG16神经网络。过程中，利用风机的振动加速度信号，通过自适应分解方法对信号进行分解降噪，将处理后信号进行变换生成二维时频谱，利用VGG16模型对时频谱样本集进行分类以达到诊断目的。为了评估该方法的有效性，将本文方法分别与传统模态分解、CNN模型进行对比，并对模型分类效果进行评估，结果表明，该方法具有更好的准确性。
关键词：风力发电机；波形分解；时频谱；VGG16
中图分类号：TK83 文献标志码：A doi：10.3969/j.issn.1006-0316.2023.10.004
文章编号：1006-0316 (2023) 10-0020-08
A Method for Wind Turbine Condition Diagnosis Based on Adaptive Decomposition and Time-Frequency Transform
HUANG Ziheng1，XU Zhaoyuan1，FANG Hui1，WU Jianbo1，LI Jinhang2，SHI Zhiyuan2
( 1.School of Mechanical Engineering, Sichuan University, Chengdu 610065, China; 2.The Research and Development Center, Dongfang Electric Corporation, Chengdu 611731, China )
Abstract：At present, the research on the working condition diagnosis method of wind turbines is relatively lacking. In this paper, a diagnosis method based on adaptive decomposition and time-frequency transform is proposed. The algorithm of this method adopts the sliding interval sample expansion, the adaptive waveform decomposition method based on improved CEEMDAN, the Hilbert transform, and the VGG16 neural network. During the process, firstly, the vibration acceleration signal is decomposed and denoised by the adaptive decomposition method, the processed signal is transformed to generate a two-dimensional time spectrum, and the VGG16 model is used to classify the time spectrum set. In order to evaluate the effectiveness of the method, the method in this paper is compared with the traditional decomposition mode and CNN model. The performance of the classification is evaluated and the result shows that the method has better accuracy.
Key words：wind turbine；waveform decomposition；time spectrum；VGG16
———————————————
收稿日期：2022-09-23
基金项目：四川省科技计划重点研发项目（2021YFG0039）
作者简介：黄子恒（1996－），男，四川成都人，硕士研究生，主要研究方向为智能控制与故障诊断，E-mail：1738568615@qq.com。*通讯作者：伍剑波（1986－），男，四川眉山人，博士，教授、博士研究生导师，主要研究方向为智能控制与无损检测，E-mail：wujianbo@scu.edu.cn。
- 上一篇：某引进型发动机喘振故障研究
- 下一篇：放射性圆柱样品几何测量气动夹具优化设计
设为首页 | 加入收藏 | 免责条款
《机械》杂志版权所有 Copyright©2008-2012 Jixiezazhi.com All Rights Reserved
电话：028-85925070 传真：028-85925073 E-mail：jixie@vip.163.com
地址：四川省成都锦江工业开发区墨香路48号 邮编：610063
Powered by PageAdmin CMS