This paper proposes a hybrid approach combining Wigner-Ville distribution (WVD) with convolutional neural network (CNN) for power quality disturbance (PQD) classification. Firstly, a WVD technique is developed to transfer a 1D voltage disturbance signal into a 2D image file, followed by a CNN model developed for the image classification. Then, the feature maps are extracted automatically from the image file and different patterns are extracted from variables on CNN. A set of synthetic signals, as well as real-world measurement data, are used to test the proposed method. The high classification accuracy of test results is achieved to confirm the effectiveness of the proposed method. Furthermore, the model is simplified and optimized by visualizing the output of convolutional layers. On this basis, one visualizing technique called the class activation map (CAM) is used to identify the location and shape of “hotspots (PQDs)”. The effect of incorrect classification of the model is analyzed with the CAM. Therefore, the proposed method is proved to have the capability of providing necessary and accurate information for PQDs, which will then be used to determine the subsequent PQ remedy actions accordingly.

This paper presents a hybrid approach combining phase space reconstruction (PSR) with a convolutional neural network (CNN) for power quality disturbance (PQD) classification. Firstly, a PSR technique is developed to transform a 1D voltage disturbance signal into a 2D image file. Then, a CNN model is developed for the image classification. The feature maps are extracted automatically from the image file and different patterns are derived from variables in CNN. A set of synthetic signals, as well as operational measurements, are used to validate the proposed method. Moreover, the test results are also compared with existing methods, including empirical mode decomposition (EMD) with balanced neural tree (BNT), S-transform (ST) with neural network (NN) and decision tree (DT), hybrid ST with DT, adaptive linear neuron (ADALINE) with feedforward neural network (FFNN), and variational mode decomposition (VMD) with deep stochastic configuration network (DSCN). Based on deep learning algorithms, the proposed method is capable of providing more accurate results without any human intervention for PQDs. It also enables the planning of PQ remedy actions.

The heat transfer of power devices has the characteristics of multi-timescale. However, the traditional thermal network model is difficult to predict the temperature information of power devices at multi-timescale accurately. This paper proposes a multi-timescale thermal network model for power devices, with MOSFET as an example. First, we establish a finite element model for temperature calculation of power devices and then reduce the order of the finite element model based on the proper orthogonal decomposition (POD) algorithm. The reduced order model is converted into an equivalent circuit model by the node voltage method and integrated into the circuit simulation software. To demonstrate its general applicability, this paper also establishes thermal network models for both IGBT and SiC MOSFET devices, validates the multi-time scale thermal network models through ANSYS/Transient Thermal software and experiments. The temperature calculation results of the three power devices all indicate that the proposed multi-time scale thermal network model can calculate the temperature of power devices faster than ANSYS/Transient Thermal model, and has less than 5% error under test conditions. This paper is accompanied by a video demonstrating a comparison of the computational speed of the multi-timescale thermal network model with a finite element model. Finally, a case study of junction temperature calculation in a Buck converter is presented to illustrate the application method of the thermal network model.

This letter presents an active feedback microwave sensor with a wide dynamic range and high sensitivity for detecting angular displacement. The sensor comprises two types of movable ring resonators (the rotor), split-ring resonator (SRR) and complement SRR (CSRR), and a modified convex-shaped open complementary split-ring resonator (the stator). The stator is designed with an active circuitry to enhance the quality factor over 4.7 K, and the feedback loop surrounds the sensing zone to enhance the mutual coupling for tracking a finite rotation of rotors in a wide dynamic range. The stator uses the separation between two reflection coefficients or the decibel magnitude variation at a fixed frequency, which is generated from its resonance frequency shift and whose values vary as a function of the rotation angle to be measured. After rotor loading, the average decibel sensitivity in terms of variation in reflection was calculated as 0.32 and 0.13 dB/degree for the SRR and CSRR, respectively, within a dynamic range of 0°–90°.

Typically, a geared drive system is used to connect an induction motor of 1500 rpm with a Raymond Pulverizer of 105 rpm in mining applications. This system suffers from low efficiency and a heavy motor drive. This paper proposes a novel design of a 75 kW, 72/48 switched reluctance motor (SRM) for a low-speed direct-drive as for mining applications. The paper is focused on the design and comparative evaluation of the proposed machine in order to replace a geared drive system whilst providing a high torque low-speed and direct-drive solution. The machine performance is studied and the switching angle configuration of the machine is also optimised. The efficiency of the whole drive system is found to be as high as 90.19%, whereas the geared induction motor drive provides only an efficiency of 59.32% under similar operating conditions. An SRM prototype was built and experimentally tested. Simulation and experimental results show that the drive system has better performance to substitute the induction motor option in mining applications.

This paper proposes a novel, two-stage and hybrid approach based on variational mode decomposition (VMD) and the deep stochastic configuration network (DSCN) for power quality (PQ) disturbances detection and classification in power systems. Firstly, a VMD technique is applied to discriminate between stationary and non-stationary PQ events. Secondly, the key parameters of VMD are determined as per different types of disturbance. Three statistical features (mean, variance, and kurtosis) are extracted from the instantaneous amplitude (IA) of the decomposed modes. The DSCN model is then developed to classify PQ disturbances based on these features. The proposed approach is validated by analytical results and actual measurements. Moreover, it is also compared with existing methods including wavelet network, fuzzy and S-transform (ST), adaptive linear neuron (ADALINE) and feedforward neural network (FFNN). Test results have proved that the proposed method is capable of providing necessary and accurate information for PQ disturbances in order to plan PQ remedy actions accordingly.