A Very-Low-Frequency Electromagnetic Inductive Sensor System for Workpiece Recognition Using the Magnetic Polarizability Tensor

A novel feature extraction method of eddy current testing for defect detection based on machine learning

“…The EM instrument operates in the spectrum from 5 kHz to 200 kHz and is capable of digital demodulation at a rate of 100 k samples per second and has an Ethernet link to a PC. It has been successfully applied to electromagnetic sensing system such as magnetic induction tomography [33, 27] and workpiece recognition [28]. A system block diagram of the EM instrument is shown in Figure 7.…”

Eco-Environmental Civilization Construction System in Remote Areas Based on Multiple Data Collection and the Internet of Things

“…Node monitoring system and edge computing network system architecture, this paper constructs a three-tier container risk monitoring model from the cloud to the edge to the user side, as shown in Figure 1. This article makes full use of network resources and meets all user needs, achieving the goal of the least total deployment cost [22,23]. Different from the service function chain deployment problem under the traditional single platform, the underlying network elements of the heterogeneous NFV environment are diversified.…”

Robust and Efficient Classification for Underground Metal Target Using Dimensionality Reduction and Machine Learning

“…The most frequently used physics-based model is the orthogonal dipole model [21], [43], [44], which is used to simulate the actual TEM response of underground metal targets in this study. When the distance from the detector to a metal target is larger than the size of the metal target, the secondary field B S can be approximated by the magnetic field generated by a dipole m [47]. In Fig.1, the secondary field B S at the position of receiving coil is calculated as:…”