关键词: 齿轮箱, 复合故障, 循环谱分析, 最大二阶循环平稳盲解卷积, 1.5维导数谱

Abstract: Aiming at the problems that the gearbox composite fault vibration signals were susceptible to severe interference from background noise, which made it difficult to accurately separate the impact characteristics of composite faults with traditional methods, an adaptive CYCBD and 1.5-dimensional derivative spectrum were proposed for composite fault diagnosis. First, the cyclic frequency components associated with the fault characteristics in the composite fault vibration signals were detected by using cyclic spectrum analysis to construct cyclic frequency sets of different target types. After that, an improved algorithm for adaptively selecting the optimal filter length of CYCBD with the third-order cumulative sparsity(TCS) as the index was proposed according to different types of cyclic frequency sets, so as to better obtain the optimal filter length containing different fault shock components. Finally, a new 1.5-dimensional derivative spectrum was proposed for feature enhancement to improve the signal-to-noise ratio and to analyze the prominent fault feature frequencies in the spectrum to discriminate the fault types. The algorithm was validated by simulated signals and fault experimental platform data, and the results show that the method may achieve the accurate separation and diagnosis of compound faults in gearboxes.

Key words: gearbox, compound fault, cyclic spectrum analysis, maximum second-order cyclostationarity blind deconvolution(CYCBD), 1.5-dimensional derivative spectrum