期刊论文

中文

计算机仿真

1006-9348

2014

31

11

88-91,99

10.3969/j.issn.1006-9348.2014.11.020

本校为第一机构

无人机在飞行的过程中,采集的故障信号数据会随着飞行环境的随机性变化而改变,造成异常信号数据关联性降低。利用传统算法进行异常信号数据挖掘的过程中,建立的关联规则过于复杂,加入过多约束条件分析环境特征,导致挖掘过程耗时较长。为解决上述问题,提出基于知识本体提纯的无人机采集实时数据挖掘方法。针对无人机异常信号数据知识本体表示方法进行了详细阐述。利用贝叶斯概率方法计算当前状态信号属于某类别异常信号的最大概率,从而为异常信号的提纯提供了准确依据。针对相似度较高的两个异常信号进行提纯处理,在提纯过程中充分考虑到信号采集时间间隔和时延,从而保证了实时性要求。实验结果表...

In the flight process of unmanned aerial vehicle （UAV） , the acquired fault signal data will change with the randomness alteration of flight environment, which reduces the correlation of abnormal signal data. In the paper, we proposed a method for the UA~ acquired real-time data mining based on the knowledge ontology refinement. The knowledge ontology representation method of UAV abnormal signal data was elaborated in this paper. Bayesian probability method was used to calculate the maximum probability that current state signals belongs to a ...