Low-voltage centralized meter reading fault detection method and device

The invention discloses a low-voltage centralized meter reading fault detection method and device, and the method comprises the steps: building a fault section feature vector according to a fault of acentralized meter reading system, and building a training set through the corresponding relation between a certain number of selected fault section feature vectors and a fault cause; training the CAEs model through the training set to obtain a CAEs fault deep feature vector extraction model, and inputting the training set fault deep feature vector into an XGboost for training to obtain an XGboostfault diagnosis model; and then extracting a real-time fault section feature vector of the system through a trained CAEs fault deep feature vector extraction model to obtain a fault deep feature vector, and inputting the fault deep feature vector into a decision tree model XGboost fault diagnosis model with stable performance and strong generalization ability to carry out fault identification toobtain a fault reason of th