Abstract:The fault diagnosis method based on vibration signal analysis depends on efficient extraction of mechanical state feature. Blind source separation is an efficient method for feature extraction, which can separate source signals from complex signals. The founded multichannel blind deconvolution model and mechanical system vibration model respectively pointed out the consistency berween these two models, verified the feasibility of using blind deconvolution to separate mechanical vibration signal. MBLMS arithmetic was used to separate vibration signals of diesel engine. The signals of piston slap, combustion and closure of input value were successfully separated, the results verified blind deconvolution can sucessfully separation desiel vibration signal.
蔡艳平;李艾华;石林锁;秦 亮;张亚洲. 基于盲解卷积的柴油机振动信号分离研究[J]. , 2010, 29(9): 38-41.
CAI Yan-ping;LI Ai-hua;SHI Lin-suo;QIN Liang;ZHANG Ya-zhou. Research on separation of diesel vibration signal based on blind deconvolution. , 2010, 29(9): 38-41.