基于猎人猎物优化算法改进粒子滤波的滚动轴承剩余使用寿命预测技术

Prediction of the remaining useful life of a rolling bearing based on a hunter-prey optimization-particle filter

针对目前滚动轴承剩余使用寿命(RUL)预测存在的预测精度不高、预测模型累计误差大等问题,提出一种融合Hodrick-Prescott(HP)趋势滤波边界线(HPTF-BL)、猎人猎物优化算法改进粒子滤波(HPO-PF)的滚动轴承剩余使用寿命预测方法。该方法首先将HP趋势滤波与退化边界线构建相结合,对表达轴承退化信息的特征指标进行处理,得到上下退化边界与主要退化趋势,然后利用猎人猎物优化算法(HPO)改进粒子滤波(PF)的重采样过程,再使用改进的方法对特征指标进行趋势预测,最后结合设定的失效阈值线得到最终的剩余使用寿命。使用美国辛辛那提大学智能维护系统(IMS)中心的轴承实验数据验证了所提方法的可行性与有效性。

Current methods for predicting the remaining useful life(RUL)of rolling bearings suffer from problems of low prediction accuracy and large cumulative errors.In this paper,we propose a method for predicting the remaining useful life of rolling bearings that combines a Hodrick-Prescott trend filter boundary line(HPTF-BL)and a hunter-prey optimization-particle filter(HPO-PF).HP trend filtering is first combined with the construction of degradation boundary lines to process the characteristic indicators that express the bearing degradation information to obtain the upper and lower degradation boundaries and the main degradation trends.The resampling process of PF is then improved by using HPO,and the trend of the characteristic indicators is predicted by using the improved method.Finally,the set failure threshold line is combined,to give the final remaining useful life.Experimental bearing data from the IMS Center verify the proposed method's feasibility and effectiveness.