【24h】

EM BASED ESTIMATION FOR HYBRID PROGNOSTIC MODELS

机译:基于EM的混合预测模型估计。

获取原文
获取原文并翻译 | 示例

摘要

Prognostic health monitoring (PHM) is an important element of condition based maintenance and logistics support. The accuracy of prediction and the associated confidence in the prediction greatly influences overall performance and subsequent actions either for maintenance or logistics support. A prognostic model is a mathematical framework for making such predictions. Accuracy of prognosis is directly dependent on how closely one can capture the system and component interactions. In this paper, we consider a hybrid model for the prognosis of deteriorating systems and come up with identification and estimation algorithms. The model considered herewith represents the degrading system as collection of prognostic states (health vectors) which evolve continuously over time. The model includes an age dependent deterioration distribution, component interactions, as well as effects of discrete events arising from line maintenance actions and or abrupt faults. Authors develop the estimation and system identification scheme for such models. This model provides non-trivial challenges for system identification and parameter estimation. In this paper, we derive Expectation Maximization (EM) based system identification and a recursive Bayesian state estimation for predicting the health of degrading asset. The efficiency of the health prediction has been demonstrated for the prognosis of APU using simulated and field data.
机译:预后健康监测(PHM)是基于状况的维护和后勤支持的重要组成部分。预测的准确性和相关的预测信心会极大地影响整体性能以及后续维护或后勤支持措施。预后模型是进行此类预测的数学框架。预后的准确性直接取决于人们能否捕捉到系统和组件之间的相互作用。在本文中,我们考虑了退化系统的混合模型,并提出了识别和估计算法。在此考虑的模型将退化系统表示为随时间连续发展的预后状态(健康向量)的集合。该模型包括与年龄相关的劣化分布,组件相互作用以及生产线维护措施和/或突发故障引起的离散事件的影响。作者开发了此类模型的估计和系统识别方案。该模型为系统识别和参数估计带来了不小的挑战。在本文中,我们导出了基于期望最大化(EM)的系统识别和递归贝叶斯状态估计,以预测降解资产的健康状况。已经通过模拟和现场数据证明了健康预测对APU预后的有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号