The implementation of power line communications (PLC) in smart electricity grids provides us with exciting opportunities for real-time cable monitoring. In particular, effective fault classification and estimation methods employing machine learning (ML) models have been proposed in the recent past. Often, the research works presenting PLC for ML-aided cable diagnostics are based on the study of synthetically generated channel data. In this work, we validate ML-aided diagnostics by integrating measured channels. Specifically, we consider the concatenation of clustering as a data pre-processing procedure and principal component analysis (PCA)-based dimension reduction for cable anomaly detection. Clustering and PCA are trained with measurement data when the PLC network is working under healthy conditions. A possible cable anomaly is then identified from the analysis of the PCA reconstruction error for a test sample. For the numerical evaluation of our scheme, we apply an experimental setup in which we introduce degradations to power cables. Our results show that the proposed anomaly detector is able to identify a cable degradation with high detection accuracy and low false alarm rate.
The advancements and reliance on digital data necessitates dependence on information technology. The growing amount of digital data and their availability over the Internet have given rise to the problem of information security. With the increase in connectivity among devices and…
The dramatic increase in the computational facilities integrated with the explainable machine learning algorithms allows us to do fast intrusion detection and prevention at border areas using Wireless Sensor Networks (WSNs). This study proposed a novel approach to accurately pred…
Electroencephalography (EEG) is a non-invasive method employed to discern human behaviors by monitoring the neurological responses during cognitive and motor tasks. Machine learning (ML) represents a promising tool for the recognition of human activities (HAR), and eXplainable ar…
Improvement of wireless power transfer (WPT) systems is necessary to tackle issues of power transfer efficiency, high costs due to sensor and communication requirements between the transmitter (Tx) and receiver (Rx), and maintenance problems. Analytical techniques and hardware-ba…
The three Ground Reaction Force (GRF) components can be estimated using pressure insole sensors. In this paper, we compare the accuracy of estimating GRF components for both feet using six methods: three Deep Learning (DL) methods (Artificial Neural Network, Long Short-Term Memor…
Due to the advancement of science and technology, modern cars are highly technical, more activity occurs inside the car and driving is faster; however, statistics show that the number of road fatalities have increased in recent years because of drivers’ unsafe behaviors. Therefor…