Research on Fault Detection Technology of Air-to-Air Missile Based on Data Mining

2022 
According to the research experience of air-to-air missile testing, the occurences of some product faults are of low frequency and sporadic, and the fault features usually hide in the details of testing signals. However, the conventional rule-based fault detection methods are difficult to issue early warning against these product faults. In order to solve the problems of miss-detection faced by the conventional missile fault detection methods, this paper started from studying the data feature extraction, feature space construction, and automatic data mining, and carried out the research on engineering application of the data-driven fault detection technology. In view of the complexity of missile test data, the feature extraction of missile test data was achieved with the applications of mathematical statistics, clustering, artificial intelligence, image feature extraction, defining a method of missile fault detection based on the \(3\sigma\) criterion, and a data-mining based fault diagnosis system was developed with the ability to evaluate product quality on the foundation of historical missile test data.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []