基于K-S检验和动态灰色模型的机械设备剩余寿命预测方法  

Remaining Life Prediction Method of Mechanical Equipment Based on K-S Test and Dynamic Grey Model

作　　者：王恒[1] 马海波[1] 徐海黎[1] 花国然[1] 

机构地区：[1]南通大学机械工程学院,江苏南通226019

出　　处：《仪表技术与传感器》2015年第1期97-100,共4页Instrument Technique and Sensor

基　　金：江苏省自然科学基金资助项目(BK2011391);南通市应用研究计划资助项目(BK2012020;BK2013026)

摘　　要：提出了一种基于K-S检验和动态灰色模型的机械设备剩余寿命预测方法。提出以Kolmogorov-Smirnov检验为基础的K-S距离作为描述机械设备退化状态的性能指标,通过退化指标序列动态训练灰色模型、更新模型参数,预测退化指标的变化趋势并确定到达设定失效阈值时的预测步数,以此计算机械设备的剩余使用寿命。最后通过轴承全寿命样本数据对其验证,并与传统的二次曲线拟合预测法和静态灰色模型预测法进行比较,结果表明所提出的方法更能有效地预测轴承的剩余寿命,具有较高的预测精度。Based on the K-S test and dynamic gray model,a remaining life prediction method for mechanical equipment was proposed. Degradation index was described by K-S distance which was based on Kolmogorov-Smirnov,and degradation index sequence was used for training dynamic gray model. Then by updating model parameters,predicting the trends of degradation indices and determining the number of steps that reach the set failure threshold,the remaining life of mechanical equipment was calculated. Finally the proposed method was tested and verified with whole life experiment data of rolling bearing and it was compared with quadratic curve fitting prediction and static gray model prediction. The results show that this method can better predict the remaining life of bearing and has higher prediction accuracy.

关 键 词：剩余寿命预测 Kolmogorov-Smirnov检验 灰色模型 轴承 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置] TH165.3[自动化与计算机技术—控制科学与工程]