【摘要】 采用KPCA进行特征提取时,其核函数参数对轴承故障特征的可分性影响很大,直接影响轴承故障诊断的准确率,而使最优的核参数难以选取。针对这一问题,采用人工蜂群算法与KPCA相结合,提出了基于人工蜂群算法的KPCA核参数优化选取方法,并实现了滚动轴承故障特征的优化提取。试验结果表明,该方法能够最大程度提高故障样本的可分性和SVM模型的分类精度;同时通过对比分析遗传算法、粒子群算法及人工蜂群算法的优化结果,验证了该方法具有更好的寻优能力。更多还原

【Abstract】 When using KPCA to extract feature,the kernel function parameters have a great influence on the separability of bearing fault feature,and affect the accuracy of the bearing fault diagnosis,thus the optimal parameters of kernel are difficult to be selected.Aimed at this problem,the artificial bee colony algorithm and KPCA were combined,then the method of KPCA kernel parameter optimization based on artificial bee colony algorithm was proposed,and it was applied in achieving the optimal extraction of rolling bearings ’ fault feature.The experimental results show that this method can improve the separability of fault features and the classification accuracy of SVM model in a considerable degree.By comparing and analysis the results of artificial colony algorithm optimization,genetic algorithm and particle swarm optimization algorithm,the conclusions verified that the proposed method in this paper has better optimization ability.更多还原