中国机械工程 ›› 2015, Vol. 26 ›› Issue (23): 3208-3213.

• 机械基础工程 • 上一篇    下一篇

旋转机械运行稳定性劣化的1.5维谱特征提取方法

蒋章雷;徐小力   

  1. 北京信息科技大学现代测控技术教育部重点实验室,北京,100192
  • 出版日期:2015-12-10 发布日期:2015-12-04
  • 基金资助:
    国家自然科学基金资助项目(51275052);北京市自然科学基金资助重点项目(3131002);北京市教委科研计划资助重点项目(KZ201311232036);现代测控技术教育部重点实验室开放课题(KF20141123202) 

Feature Extraction Method of 1.5-dimensional Faced to Running Stability Deterioration of Rotating Machinery

Jiang Zhanglei;Xu Xiaoli   

  1. Key Laboratory of Modern Measurement & Control Technology,Ministry of Education,Beijing Information Science and Technology University,Beijing,100192
  • Online:2015-12-10 Published:2015-12-04
  • Supported by:

摘要:

风电机组是一种典型的大型旋转机械,其运行状态具有变工况、非平稳特点,运行中工况和负载等非故障因素的变化会造成信号能量产生变化,因此传统的基于能量的振动级值及功率谱方法难以有效实现运行稳定性劣化特征的提取。针对该情况,研究了高阶累积量运行稳定性劣化特征提取方法,提出了基于敏感性、趋势性、差异性、一致性判断特征提取方法的趋势预测适用性的方法,通过转子实验台多种劣化类型下不同劣化程度状态的实验研究,检验了1.5维谱方法对状态劣化的敏感性、趋势性、差异性、一致性,评估了该方法作为特征提取手段的性能,解决了风电机组传动系统运行稳定性劣化的状态诊断、劣化趋势预测中特征提取方法的选择缺少理论依据的问题。

关键词: 传动系统, 稳定性劣化, 1.5维谱, 特征提取

Abstract:

Wind turbines were typical large rotating machinery with variable conditions and non-stationary states, and vibration signal energy was changed because of non-fault factor such as variable conditions and loads. Fault diagnosis method based on vibration signal energy such as vibration level and power spectrum method could not extract running stability deterioration features from non-stationary signals under the influences of variable conditions and loads. 1.5-dimensional spectrum feature extraction method was studied based on higher-order cumulant method,and an evaluation method was proposed by using sensitivity property, trend property, difference property, and consistency property. Experimental data of varying degrees of deterioration under various types of deterioration were carried out to validate the proposed method, and the results show that feature extracting method can separate the deterioration characteristics from non-degradation characteristics, and the evaluation method can provide theoretical basis for selecting feature extraction method for fault diagnosis and trend prediction of running stability deterioration.

Key words: transmission system, stability deterioration, 1.5-dimensional spectrum, feature extraction

中图分类号: