Machine learning or deep learning methods, e.g. deep neural networks (DNNs), support vector machine (SVM), random forest (RF), et al. can be used to solve classification or regression problems which exist in the intelligent maintenance of power system. Some new AI technology, e.g. federated learning (FL), can also be applied to assist power maintenance with obeying data privacy regulations and maximize the value of data.
Recognize towers of transmission lines using convolution neural networks (CNNs)
Squeeze-and-Excitation improved fast unsupervised anomaly detection with generative adversarial network (SE-f-AnoGAN) for power tower anomaly detection from UAV inpection images
Recognize towers of transmission in mountain areas
Federated learning (FL)-based visual classification and detection for processing power inspection images
Federated learning (FL)-based visual classification and detection
Fault diagnosis of on-load tap changers (OLTCs) based on improved semi-supervised ladder networks
Structure of three-layer Ladder Networks for processing vibration signals of OLTCs
Experiment result of Ladder Networks for fault diagnosis of OLTCs
Predict condition of gas-insulated switchgears (GIS) using support vector machine (SVM)
Predict openning travel curves of GIS