In this paper, a set of dissolved gas analysis (DGA) new feature combinations is selected as input from the mixed DGA feature quantity, and an improved krill herd (IKH) algorithm optimized support vector machine (SVM) transformer fault diagnosis model is established to solve the problem that the single characteristic gas or characteristic gas ratio, which are utilized as the DGA feature quantity cannot fully reflect the transformer fault classification. The following work has been done in this paper: 1) IEC TC 10 fault data and other 117 sets of fault data in China are preprocessed in order to reduce the influence on the diagnosis results causing by the edge data in the fuzzy area; 2) the SVM parameters and 11 features are encoded by a binary code technique; 3) a preferred DGA feature set for fault diagnosis of power transformers is selected by genetic algorithm (GA) and SVM, and; 4) IKH is utilized to optimize the parameters of SVM. Combining with cross-validation principle, a transformer fault diagnosis model based on IKH algorithm to optimize SVM is established. The fault diagnosis results based on the new fault sample show that the proposed DGA feature set to increase the accuracy by 26.78% and 10.83% over the DGA full data and IEC ratios. Moreover, the accuracy of IKHSVM is better than the GASVM, back-propagation neural network (BPNN), and particle swarm optimization optimized support vector machine (PSOSVM), the accuracy rates are 85.71%, 75%, 64.29%, and 71.43%, which proves the validity of the proposed fault diagnosis model. INDEX TERMS Power transformers, fault diagnosis, support vector machine, improved krill herd algorithm, DGA feature.

No abstract

The increasing household loads make series arc faults more complex, which are difficult to be detected by traditional circuit breakers and lead to the frequent occurrence of residential fire accidents. In this paper, a comprehensive approach of complex load recognition and series arc detection is proposed on the basis of principal component analysis and support vector machine (PCA-SVM) combination model. Several typical loads were selected and analyzed, especially nonlinear and complex loads like power electronics load and multi-state load. Three time-domain parameters, maximum slip difference (MSD), zero current period (ZCP) and maximum Euclidean distance (MED), and nine frequency-domain harmonics information are collected to complex waveforms. To decrease the computation cost and further to enhance the response velocity, all the time-domain and frequency-domain information were blended and dimensionally reduced to three parameters by principal component analysis (PCA). Prior to the series arc detection, load recognition is trained and completed with the artificial intelligence (AI) algorithm. At last, the comprehensive model of load recognition and series arc detection is achieved based on a support vector machine (SVM). The accuracy of load recognition and series arc detection reaches 99.1% and 99.3%, respectively, demonstrating the excellent performances of the intelligent approach to diagnose the series arcing activities in modern household applications.

No abstract

By using a ballast resistor having resistance of 10 MΩ, varying the gap distance from 23 mm to 0 mm and using a fixed dc voltage at 14 kV, the streamer corona, single filament, transient glow, dc glow, and spark modes and their transitions are demonstrated in positive needle-to-plate air discharge at atmospheric pressure. The electrical characteristics, the rotational temperature, and vibrational temperature of N2, as well as the temporal behavior of streamer propagation in these discharge modes, are investigated. First, to the best of our knowledge, the transient glow mode between the single filament mode and the dc glow mode, operated in a stable repetitive fashion, is reported for the first time in positive dc air discharges. The pulse repetition frequency ranges from 7.5 to 15 kHz. The current density and the rotational temperature are in the range of 27–105 A/mm2 and 600–850 K, respectively. Its temporal behavior reveals that after the primary streamer arrives at the cathode, the secondary streamer initiates within several nanoseconds near the anode and then propagates at a high speed of 105–106 m/s. There is no transition to spark even after the secondary streamer arrives at the cathode. Second, the transition from single filament to transient glow is characterized by the sudden decrease in the pulse repetition frequency and the abrupt increase in the current amplitude, the pulse width, and the gas temperature. Third, the transition from transient glow to dc glow is identified visibly by the formation of typical glow structure (positive column, Faraday dark space, and negative glow), which is accompanied by the transition of the discharge current from nanosecond pulse to dc. In addition, both the ballast resistor and the stray capacitor exert significant influence on the transition of discharge modes.

Bushings are served as an important component of the power transformers; it's of great significance to keep the bushings in good insulation condition. The infrared images of the bushing are proposed to diagnose the fault with the combination of image segmentation and deep learning, including object detection, fault region extraction, and fault diagnosis. By building an object detection system with the frame of Mask Region convolutional neural network (CNN), the bushing frame can be exactly extracted. To distinguish the fault region of bushings and the background, a simple linear iterative clustering‐based pulse coupled neural network is proposed to improve the fault region segmentation performance. Then, two infrared image feature parameters, the relative position and area, are explored to classify fault type effectively based on the K‐means cluster technique. With the proposed joint algorithm on bushing infrared images, the accuracy reaches 98%, compared with 44% by the conventional CNN classification method. The integrated algorithm provides a feasible and advantageous solution for the field application of bushing image‐based diagnosis.

Non-thermal plasma catalysis, as a special heterogeneous catalytic reaction, needs to consider both gas discharge and catalytic reaction. Packed bed dielectric barrier discharge (PB-DBD) is widely used in non-thermal plasma catalysis, but the exact control principle of gas discharge, especially streamer discharge, is not clear. In this study, therefore, the orderly arranged dielectric rods were packed in the discharge gap of PB-DBD, and the streamer discharge behaviors were controlled by adjusting their diameter(s), quantity(ies), location(s) and dielectric constant(s). Al2O3 and ZrO2 dielectric rods with dielectric constants of about 9 and 25 were used as packing material. Pure CO2 was used as reaction gas and discharge gas. Discharge images showed that stable and controllable streamer discharges can be formed between the dielectric rod and ground electrode. The intensity, width and length of the streamer discharge can be significantly changed by optimizing the dielectric constant, diameter, packing number and position of the dielectric rod, thereby affecting the CO2 conversion efficiency. Increasing dielectric constant and the distance between the dielectric rod and ground electrode can increase the intensity of streamer discharge, thus promoting the CO2 conversion efficiency. Compared with an empty reactor, after packing 24 ZrO2 dielectric rods with a diameter of 1 mm, the CO2 conversion and energy efficiency increased from 9.58% to 20.1% and from 1.67% to 2.89%, respectively. In short, this research has important implications for plasma catalysis. This study not only reveals the synergistic characteristics between streamer discharge and CO2 dissociation, but also provides an important idea for structural optimization of PB-DBD catalyst.

With plasma penetration into a capillary and the assistant of an air DBD, an ac‐driven Ar microplasma plume having a length of several cm is generated inside the capillary. The inner diameter of the capillary ranges from 4 to 100 µm. When the tube diameter decreases, the length of the microplasma plume decreases, and while the ignition voltage, the current density, and the electron density increase significantly. For the tube diameter of 9 µm, the current density and the electron density measured from the Stark broadening of Hα line reach as high as 109 A m−2 and 11 × 1016 cm−3, respectively. The microplasma plume is of high degree of ionization and remains non‐thermal. Rather than monotonically increasing, the propagation velocity of the microplasma plume decreases and then keeps almost unchanged after the tube diameter reaches 20 µm.