机械系统实测信号预处理方法研究现状与展望

Status and prospect of research in preprocessing methods for measured signals in mechanical systems

在机械系统状态监测和故障诊断中,对实测信号进行预处理是一个必要环节,是保证后续特征提取和健康状态评估可靠性的基础.综述了近年来国内外学者在机械系统实测信号预处理方法研究方面的探究成果,介绍了信号类型转换方法的研究现状,回顾了信号有效成分提取方法的研究成果,对机械系统实测信号预处理发展趋势进行了分析与展望.指出了关于传统时域、频域积分方法的改进和性能提升,尚存在许多工作需要进一步深入探究;采取智能信息处理方法获得能够反映设备状态特征的完备数据源,仍然是值得深入研究的方向.分析表明,随着机械装备逐步趋向于智能化、多功能化、高效化和绿色化,机械系统实测信号预处理技术将迎来新的挑战;探究先进的实测信号预处理方法,依据设备运行状态信息,准确进行智能故障诊断和故障预测,在科学研究与工程实际中具有广阔的应用前景.

In condition monitoring and fault diagnosis of mechanical systems,the preprocessing of measured signal is a necessary link and a basis ensuring subsequent feature extraction and reliability of health assessment.The results achieved by existing preprocessing methods for measured signals were reviewed.The status of signal conversion methods was described,and the outcomes of signal component extraction methods were introduced.Moreover,the development trend of measured signal preprocessing in mechanical systems was analyzed and prospected.Much work in update of existing traditional time-domain and frequency-domain integration methods and their performance improvement still needs to be exploited indepth.Intelligent processing methods can obtain complete data source,which reflects the state characteristics of equipment,and are still worthy being investigated further.As the mechanical equipment gradually tends to be intelligent,multi-functional,the efficient and green the preprocessing technology for measured signals in mechanical systems will face new challenges.Advanced signal preprocessing methods should be explored to carry out precise intelligent fault diagnosis and prediction based on information of equipment operating state.Such methods can find extensive applications in scientific research and engineering practice in the future.