High-voltage circuit breaker unmarked fault diagnosis method

The invention relates to the field of high-voltage circuit breaker fault diagnosis, in particular to an unmarked fault diagnosis method for a high-voltage circuit breaker, which comprises the following steps of: S1, extracting feature vectors of the high-voltage circuit breaker under various working conditions, and classifying the feature vectors according to working condition categories; s2, randomly selecting a feature vector set under one fault working condition as a test set, and taking the rest as a training set; s3, training a single-classification classifier according to the feature vector set in the training set; s4, introducing a model tightness parameter lambda, and verifying the single-classification classifier through the feature vectors in the test set to obtain an optimal model tightness parameter lambda; and S5, repeating the steps S2-S4 until the feature vector set under each fault working condition is selected to be used as a test set, and determining a final model tightness parameter. Accord