系统工程与电子技术

• 系统工程 • 上一篇    下一篇

基于数据深度的设备状态评估模型研究

郭忠来1,2, 吴华2, 胡永刚2, 张为华1   

  1. 1. 国防科学技术大学航天科学与工程学院, 湖南 长沙 410073;
    2. 中国酒泉卫星发射中心, 甘肃 兰州 732750
  • 出版日期:2014-05-22 发布日期:2010-01-03

Equipment state evaluation model based on data depth

GUO Zhong-lai1,2,WU Hua2,HU Yong-gang2,ZHANG Wei-hua1   

  1. 1. School of Aerospace and Materials Engineering, National University of Defense Technology, Changsha 410073, China;
    2. Jiuquan Satellite Launch Center of China, Lanzhou 732750, China
  • Online:2014-05-22 Published:2010-01-03

摘要:

针对复杂类型设备在线监测问题,提出了一个基于数据深度的设备状态评估一般模型。首先找到设备的最佳状态,通过理想点的对称变换,得到历史样本与现有样本的合集;然后利用数据深度的原理,计算当前样本与合集样本的相似度量,作为设备当前状态评价的结果。对于过程状态,即函数型数据,首先利用多元数据的方法计算出同一时刻的评价结果,然后利用沿时间轴积分的方法计算出平均结果。文中的实验结果验证了方法的有效性。

Abstract:

For online monitoring of complex equipment, a novel method for equipment state evaluation is proposed. Firstly, according to the desired point of the equipment, the central symmetrical sample set is obtained, and then the original sample and the symmetrical sample are combined as one target sample set. Secondly, the depth of the current status with respect to the target sample is taken as the performance evaluation result. Finally, for the functional data, the evaluation result at the same moment is obtained with the multivariate date method and then the average result is obtained with the integral method along the time axis. The experimental results show the validity of the methods.