Ensemble deep learning for automated classification of power quality disturbances signals

• A deep learning framework for intelligent classification of PQD signals is proposed. The improved Bagging algorithm has better accuracy and generalization performance. active strategy selects the most representative samples from a dataset. uses less labeled data thus saving manual labeling cost. automatic power quality disturbances (PQD) great significance solving problems. In this study, we propose an ensemble to realize PQ disturbances. Specifically, based on characteristics sequence disturbance signals, Long Short Term Memory (LSTM) network used classify signals. addition, theory integrate training results multiple LSTM networks improve network. Our contribution lies in combination extract representation view large number unlabeled grid, adopted select set, which can enhance model performance with data. Finally, experiments were conducted different noise environments. Compared existing multi-label models, method achieves good calculation speed. Furthermore, proposed able train fewer samples, reducing costs.