基于LCD的齿轮箱混合故障盲源分离研究

盲源分离是一种有效的混合故障诊断方法,而局部特征尺度分解（LCD）是非平稳信号的有效分析处理工具，综合两者的优点，提出了基于LCD的齿轮箱混合故障盲源分离方法。将源信号LCD分解，得到新的多维信号，采用Bayesian信息准则（BIC）估计盲源的数目并对多维信号进行重组。最后进行联合近似对角化处理，实现源信号的盲分离。仿真和实验结果表明，该方法能够有效地实现齿轮箱混合故障盲源分离。

BSS Method of Gearbox Multi-fault Based on LCD

BSS is an effective multi-fault diagnosis method, and LCD is an effective tool of analyzing and processing non-stationary signals. Combining the advantages of LCD and BSS, a BSS of gearbox multi-fault was proposed based on LCD. The source signals were  decomposed by LCD to constitute new multidimensional signals, then estimating blind source number by Bayesian information criterion(BIC) to restructure the multidimensional signals, finally, using joint approximate diagonalization to separate the blind source signals. Simulation and experiments show that the method can effectively separate the blind source of the gearbox multi-faults, and provide a new method for fault diagnosis of the gearbox.