摘要
针对蓄能机组水轮机轴承故障检测准确率不足的问题,提出对多源信息特征进行提取和检测。具体则是对NMD算法进行改进设计,以提升分解信号的准确度,保证故障诊断的准确度。改进的NMD算法中,完善了主谐波与次谐波的提取终止条件,并对信号分解流程作出了调整,使其信号分解逻辑更为清晰;基于GMP拟合进行时频脊线提取,可有效避免噪声对时频脊线提取的影响,提升后续重构信号的准确度。仿真实验证明,改进的NMD算法具有较好的信号去噪性能,且重构的振动信号的准确度较高,不存在失真的情况,基本满足改进需求;经实体实验证明,改进的NMD算法能够从含有噪声和异常尖峰信号的原始信号中,准确将振动较大的振动信号准确提取出来,在蓄能机组水轮机轴承故障检测中具有一定的应用效果。
In response to the problem of insufficient accuracy in fault detection of hydraulic turbine bearings in energy storage units,a fault detection method based on multi-source information fusion is proposed.Specifically,the NMD algorithm is improved and designed to improve the accuracy of decomposed signals and ensure the accuracy of fault diagnosis.In the improved NMD algorithm,the extraction termination conditions for main and sub harmonics have been improved,and the signal decomposition process has been adjusted to make its signal decomposition logic clearer;Extracting time-frequency ridges based on GMP fitting can effectively avoid the impact of noise on time-frequency ridge extraction and improve the accuracy of subsequent reconstructed signals.Through simulation experiments,it has been proven that the improved NMD algorithm has good signal denoising performance,and the reconstructed vibration signal has high accuracy without distortion,basically meeting the improvement requirements;Through physical experiments,it has been proven that the improved NMD algorithm can accurately extract vibration signals with high vibrations from raw signals containing noise and abnormal peak signals,and has certain application effects in fault detection of hydraulic turbine bearings in energy storage units.
作者
贾海军
沈华哲
王志楠
JIA Haijun;SHEN Huazhe;Wang Zhinan(Generation Company of China Southern Power Grid Co.,Ltd.,Guangzhou 510430,China)
出处
《自动化与仪器仪表》
2024年第4期76-81,共6页
Automation & Instrumentation
基金
2022年南网储能科技公司科技项目《蓄能机组发电机、水轮机综合状态评估方法及检修策略研究》(022200KK52190003)。
关键词
NMD算法
信号去噪
信号分解
故障诊断
NMD algorithm
signal denoising
signal decomposition
fault diagnosis
