采用EEMD-KPCA处理的IHHO-LSSVM滚动轴承寿命预测模型

为提高滚动轴承剩余寿命预测精度,提出一种基于集合经验模态分解-核主成分分析(EEMD-KPCA)和改进的哈里斯鹰优化-最小二乘支持向量机(IHHO-LSSVM)的滚动轴承剩余寿命预测模型.首先,使用集合经验模态分解方法对原信号进行分解,根据相关系数和峭度值选取合适的本征模态函数进行重构;然后,提取时域、频域、小波包能量谱等指标,并用核主成分分析,选取累计贡献率大于85%的主成分作为轴承退化性能指标;建立最小二乘支持向量机寿命预测模型,针对模型参数,提出一种改进的哈里斯鹰优化算法,并在新算法基础上设计新的能量周期性递减调控机制.采用轴承全寿命实验数据进行验证,结果表明:该方法提取的轴承性能评估指标能够更全面地表征轴承性能退化情况,建立的模型具有良好的预测效果.

IHHO-LSSVM Rolling Bearing Life Prediction Model Treated by EEMD-KPCA

In order to improve the prediction accuracy of the remaining life of the rolling bearing,a model based on ensemble empirical mode decomposition-kernel principal component analysis(EEMD-KPCA)and improved Harris hawk optimization-least squares support vector machine(IHHO-LSSVM)was proposed. First, the EEMD method was used to decompose the original signal,and the appropriate eigenmode function was selected for reconstruction according to the correlation coefficient and kurtosis value. Then, the time domain and frequency domain indicators together wavelet packet energy spectrum, etc. were extracted, and using nuclear PCA, the principal component with a cumulative contribution rate greater than 85% was selected as the bearing degradation performance index. The LSSVM life prediction model was established, the parameters of which were optimized by a proposed IHHO algorithm. On the basis of the new algorithm,a new energy cyclical decrease regulation mechanism was designed. The rolling bearing full life experimental data was used for verification. The results show that the bearing performance evaluation index extracted by the proposed method can more comprehensively characterize the degradation of rolling bearing performance,and the established model has a good prediction effect.