基于循环谱密度切片集合分析法的轴承故障诊断

Bearing Faults Diagnosis Based on Cyclic Spectral Density Slice Gathering Method

分析了循环自相关函数和循环谱密度函数的解调性能、对噪声的免疫性能及其运算量大的局限性,兼顾故障特征提取的准确性和计算量,提出了峰值频率切片集合分析法:取循环自相关切片图上的各峰值频率作为循环频率分别做循环谱密度切片图,然后对图中f域的谱峰规律进行分析。滚动轴承故障实验表明该方法能有效排除非故障频率干扰,快速诊断出弱小故障。

The performance of the CAF and CSDF which includes their demodulation ability, de-noise ability and limitations due to amount of computation was analyzed briefly. Considering both accurate on feature extraction and tolerable with computation complexity, a slices gathering method which regarding the peak frequencies of CAF slice （r=0） as cyclic frequency α and drawing a set of CSDF slices, then searching the peak points relation at frequency domain was proposed. The experimental result on beatings shows that the method can eliminate noise and diagnose faults effectively and quickly. wiling