摘要
针对一类剩余寿命预测问题,提出基于Gamma过程的状态空间退化模型用于描述装备的性能退化过程,对装备的剩余寿命进行预测.为了解决模型求解过程中,状态空间模型隐状态通常难于测量、收集的数据不完整,并且监测值不确定的实际问题,提出了经验最大化算法和粒子滤波算法结合的方法对模型参数进行求解.在案例研究中,建立了直升机主减速器行星架裂纹和振动信号特征之间的状态空间模型,进行剩余寿命预测,结果表明基于Gamma过程的状态空间退化模型能够较为合理的预测其剩余寿命.
In this paper, for residual useful life prediction, the state space model based Gamma Process is proposed to describe the degradation of equipment performance. In the process of model,it often encountered the following problems. 1）the hidden status of equipment is difficult to measure;2）the collected data is usually incomplete; 3）the monitoring values is uncertain. To solve the above problems, we propose the method combined Experience maximization algorithm and particle filter algorithm to get the model parameters.In the case studies, the model of state space between the crack of planetary gear carrier plate of helicopter and the vibration signal features is established to predict RUL.Finally, the result shows that the state space degradation model base on Gamma Process can have more reasonable prediction in URL.
出处
《军械工程学院学报》
2015年第2期1-7,共7页
Journal of Ordnance Engineering College
基金
总装备部重点预研基金资助项目(9140A27020308JB34)
