面向飞行器遥测数据的关联规则挖掘方法研究

Study on Association Rules Mining Method for Spacecraft Telemetry Data

摘要:

飞行器遥测数据是飞行器状态的直接体现，对飞行器遥测数据的不断深入分析和研究，可为飞行器的安全性和稳定性提供有效保障。目前复杂飞行器的遥测数据存在试验数据量大、人工判读效率低、数据间关联关系复杂且不易梳理等问题。同时，数据智能化分析程度低，缺少对海量历史试验数据的有效利用。为克服现有技术不足，通过对飞行器遥测数据的关联规则挖掘方法进行研究，提出基于状态转换提取的关联规则挖掘算法，并与FP-Growth算法进行试验挖掘对比分析，实现对飞行器遥测数据参数的关联规则挖掘分析，有效地解决飞行器遥测数据间关联规则的梳理问题，试验结果准确率高，为飞行器工况与参数的关联规则挖掘提供重要参考意义。

Abstract:

Spacecraft telemetry data is a direct manifestation of spacecraft status. Continuous in-depth analysis and research on spacecraft telemetry data can provide effective guarantees for the safety and stability of spacecraft. At present, the telemetry data of complex spacecraft has problems such as large amount of test data, low manual interpretation efficiency, complex associations between data, and difficult to sort out. At the same time, the level of intelligent data analysis is low, and the effective use of massive historical test data is lacking. In order to overcome the shortcomings of the existing technology, by researching the association rules mining method of spacecraft telemetry data, the association rule mining algorithm based on state transition extraction is proposed, and conduct experimental mining and comparative analysis on the FP-Growth algorithm, implement the mining and analysis of the association rules of spacecraft telemetry data parameters, and effectively solve the problem of combing the association rules among the spacecraft telemetry data. The test results are highly accurate and provide important reference for mining the association rules of spacecraft operating conditions and parameters.