基于果蝇优化算法小波支持向量数据描述的滚动轴承性能退化评估

摘要

A rolling bearing performance degradation assessment method was proposed based on FOA-WSVDD,aiming at the problems that the SVDD algorithm was not sensitive to rolling bearing early faults and difficult to select suitable kernel parameters.The feature vectors of the time and time frequency domains were extracted from bearing fault-free stages and then were selected based on monotonicity.Then,the FOA-WSVDD model was established where the wavelet kernel function was introduced to overcome the problems that the existing kernel function was not sensitive to the early faults of the rolling bearings,and kernel parameters were optimized based on the improved FOA where the ratio of the numbers of support vectors and the total samples was used as fitness function. Finally,feature vectors were input into the WSVDD model,and the bearing performance degradation index was obtained.The experimental results show that the proposed method may accurately predict the bearing early faults,and it is 1 7 hours earlier than that of the SVDD algorithm which is based on Gauss kernel function.%针对支持向量数据描述(SVDD)算法对滚动轴承早期故障不敏感、参数选择困难的问题,提出了一种基于果蝇优化算法 小波支持向量数据描述(FOA-WSVDD)的滚动轴承性能退化评估方法.提取滚动轴承早期无故障振动信号的时域、时频域特征向量,并基于单调性进行特征选择;针对现有核函数对滚动轴承早期故障不敏感问题,将小波核函数引入到 SVDD 算法中;针对 SVDD 算法参数选择困难的问题,以支持向量个数与总样本数的比值作为适应度函数,采用改进的 FOA 算法对其核参数进行优化,建立FOA-WSVDD评估模型;最后,将轴承后期振动数据的特征向量输入到该 WSVDD 模型中,得到轴承的性能退化指标.试验结果表明,采用所提方法能准确地对轴承早期故障作出预警,与基于高斯核函数的SVDD算法相比,提前了17 h.