LSTM power quality disturbance classification with wavelets and attention mechanism

Abstract

Efficient detection and classification of power quality disturbances is required with the increasing penetration of multi-energy systems such as microgrids and features from renewable energy resources. Machine learning approach is popular to generate useful and optimal features from data learning to improve the classification performance. This paper aims to analyse the classification performance using the hybrid model of multi-resolution analysis and long short-term memory network. The proposed model uses four-level decomposition wavelet transform to increase the resolution of input signals into multi-bands signal representation. Spatial and temporal feature representation of the wavelet coefficients are highlighted using attention mechanism before feeding into long short-term memory network for sequence feature extraction. The sequence feature output is then passed into multiple dense layer for the classification process. Synthetic disturbance signals are used as training samples. The performance test carried out includes the condition of 20–50 dB signal-to-noise ratio signals, where additive white Gaussian noise are added into the test samples.

Application of Methods Based on Artificial Intelligence and Optimisation in Power Engineering—Introduction to the Special Issue

“…The results showed that the CNN-LSTM method had the highest accuracy, achieving an accuracy of 98.9%. Exactly the same accuracy was achieved by the LSTM method developed by Chiam et al [105]. The method developed by Rajiv [106] demonstrated the highest accuracy among those discussed in this work.…”

“…The issue of classifying power quality disruptions through machine learning based on LSTM, which is a type of RNN, is also described in [101][102][103][104][105][106][107][108]. In [108], an LSTM method combined with a CNN method was proposed, which achieved an accuracy of 97.3%.…”