Aerostat capsule surface damage identification method

The invention discloses an aerostat capsule surface damage identification method, and belongs to the technical field of capsule surface damage identification. Aiming at the problem of aerostat capsule surface damage identification, the invention establishes an aerostat capsule surface damage identification method based on CWT and CNN, can accurately identify the damage of the aerostat capsule surface, realizes long-term monitoring of the capsule state, and ensures the safe operation of an aerostat. The adopted CWT-CNN model can realize classification of various damage states at the same time, and the problem of low classification precision of traditional machine learning is solved; the adopted CWT-CNN model is different from a traditional fault diagnosis method and needs to experience a traditional feature extraction process, and the defect that the traditional fault diagnosis method needs to depend on a large number of signal processing technologies and rich engineering practical experience to extract fault