基于激励分析的齿轮箱故障诊断研究

In this project, a new gearbox fault diagnosis theory and method based on exciting analysis is presented and studied. In theory, complex gearbox modeling analysis method which take shaft system and shell as sub-system is presented, the structural non-linear and time-variation problem are solved, and the dynamic relationship of component working condition and exciting on bearing block are analyzed; To improve identifing precision, multi-load identification methods are studied and some methods are presented for the first time： relatively error least square method in FRF reconstruction, and multi-coherence correction method and etc; analysis is made to study fault feature drawing methods based on exciting analysis, and a feature system based on exciting analysis is established; the fuzzy clustering methods are applied to detect and diagnose the gearbox and experiment shows diagnosis methods based on exciting analysis has high diagnosis precision and a nice fault orientation. On the experiment, multi-load identification bench, gearbox experiment bench are established and gearbox fault diagnosis experiment technique based on exciting analysis are studied, the multi-load identification and gearbox fault diagnosis test technique are researched, and selfpropelled gun gearbox typical fault experiment is proceeded. Experiments show this method have advantage of the high diagnosis accuracy and accurate orientation, and is a new development to gearbox fault diagnosis theories and technique.

研究基于激励分析的齿轮箱故障诊断新理论和新方法。研究内容：1建立齿轮箱系统的整体力学模型并对零部件典型故障对振动激励的影响进行分析；2研究高精度的多载荷识别理论及数据处理方法并经实验验证；3研究基于激励信号的牲撮和模式识别理论与方法，此研究将提高齿轮箱故障诊断精度，促进载荷识别理论和复杂齿轮箱故障诊断技术的发展。.