A Portable Power Quality Monitoring Approach in Microgrid with Electromagnetic Sensing and Computational Intelligence

Monitoring power quality (PQ) in microgrids is gaining increasing attention in recent years due to the popularity of microgrids and PQ disturbances caused by renewable energies. Many techniques based on artificial neural networks (ANNs) are proposed for monitoring the PQ with no need to pre-set thresholds. However, the necessity of retraining the ANN is a big problem when the electrical parameters vary. This article proposes a new approach to detect and classify the PQ disturbances accurately in multimicrogrids based on electromagnetic sensing and portability-enhanced ANN. The proposed ANN-based approach avoids the retraining of weights, when the voltage, current, and frequency varies with microgrids. Two steps are critical for achieving the portability of the ANN in various microgrids, which are pre-normalization and using the same maximum and minimum feature vectors for feature matrix normalization. Meanwhile, the electromagnetic sensing facilitates non-intrusive monitoring and easy installation. The high accuracy of simulation and experimental results in various scenarios validate the effectiveness and efficiency of this portable and non-invasive approach for monitoring PQ in multimicrogrids without retraining ANN.