SBPD算法在行程开关故障预测中的研究  

Study on the Method of State Based Prognostics with State Duration Information(SBPD) in Failure Prediction of Limit Switch

作　　者：陈长骏[1] 王凌[1] 潘静[1] 陈锡爱[1] 许宏[1] 那文波[1] 

机构地区：[1]中国计量学院机电工程学院,浙江杭州310018

出　　处：《自动化仪表》2016年第1期21-25,共5页Process Automation Instrumentation

基　　金：浙江省自然科学基金资助项目(编号:LY14F030019;LQ14F050003)

摘　　要：针对行程开关弹簧片的弹力寿命有限的故障,提出了基于状态驻留信息(SBPD)模型的故障预测方法。首先,应用均值与聚类评估指标对行程开关劣化状态进行评估和优化。其次,在对行程开关劣化状态识别的基础上,提出了基于状态驻留信息模型的剩余使用寿命计算方法。最后,在对行程开关进行加速疲劳实验的基础上,应用实验数据将SBPD模型与马尔可夫链蒙特卡罗仿真(MCMC)方法进行验证比较。结果表明,在行程开关弹簧片的弹力寿命有限的故障预测中,SBPD模型优于MCMC仿真方法,可有效预测行程开关剩余使用寿命。Aiming at the faults caused by limited stretch life of the spring leaf in travel switch, the failure prediction method using state-based prognostics with duration information （ SBPD ） model is proposed. Firstly,the evaluation and optimization of the deterioration state of the travel switch are conducted by using K - means and clustering evaluation index; then, on the basis of identification of deterioration state, the calculation method of the remaining useful life （ RUL ） based on SBPD model is proposed; finally, based on the accelerated fatigue experiment of the travel switch,the validation comparison between methods of SBPD model and Markov Chain Monte Carlo （MCMC） simulation is conducted with the experimental data. The results show that the SBPD model is better than MCMC simulation in prediction of faults caused by limited stretch life of the spring leaf; it can effectively predict the remaining useful life of the travel switch.

关 键 词：故障预测 行程开关 剩余使用寿命预测 状态驻留预测模型 马尔可夫链蒙特卡罗仿真 

分 类 号：TH17[机械工程—机械制造及自动化] TP29[自动化与计算机技术—检测技术与自动化装置]