China
Africa
Explore chapters and articles related to this topic
Special Features of SimPowerSystems Models
Published in Viktor M. Perelmuter, Electrotechnical Systems, 2020
Grounding Transformer is used in distribution networks for providing of a neutral point in a three-wire system, usually, for supplying of the single-phase loads connected to the ground (Figure 2.36). The transformer has three primary and three secondary windings connected in zigzag, and all six windings have the same number of the turns; in each phase, “belonging to it” primary and “foreign” secondary windings connected with opposite polarity (anti-aiding). The voltage of each of six windings is V/3, where V is the rated network voltage. The current distribution is shown in Figure 2.37.
Transformers
Published in T. A. Short, Electric Power Distribution Handbook, 2018
A grounding transformer must handle the unbalanced load on the circuit as well as the duty during line-to-ground faults. If the circuit has minimal unbalance, then we can drastically reduce the rating of the transformer. It only has to be rated to carry short-duration (but high-magnitude) faults, normally a 10-sec or 1-min rating is used. We can also select the impedance of the grounding transformer to limit ground-fault currents.
Earthing Transformers
Published in K.R.M. Nair, Power and Distribution Transformers, 2021
As per IEC 60076-6, earthing transformers (also termed as neutral couplers) are used to provide a neutral connection for earthing a 3-phase ungrounded network. Earthing transformer is termed grounding transformer as per ANSI C.57.12, and it is described as “a Transformer intended primarily to provide a neutral point for grounding purposes on a 3-phase ungrounded system to provide a return path for the fault current and to support a faulted phase above ground”.
Simulation and experimental research on methods to suppress ferroresonance in potential transformers
Published in Journal of Control and Decision, 2018
Hongwen Liu, Ke Wang, Junhui Zhao, Hao Li, Shijin Tian
In order to study the reliability of common anti-ferroresonance methods, a series of experiments to reproduce the ferroresonance phenomena were conducted on a 10 kV distribution network. The test system diagram is shown in Figure 9. The system contains an 800 kVA three-phase source, several 400 V, 500 A low-voltage switches, a 400 kVA voltage regulator, a grounding transformer (DKSC 3500/38.5-500/0.4) with an arc suppression coil, a 12 kV medium voltage breaker, PTs (JDZX71-10), several fuses (XRNP-12/0.5-50), several primary resonance eliminators (LXQ(D)III-10), a number of capacitors with different ratings and so on. The grounding fault is removed by a medium voltage breaker. During the tests, the waveforms of ferroresonance overvoltage and PT current were monitored and recorded by a DF1024 high-speed data acquisition card.
GMTS: GNN-based multi-scale transformer siamese network for remote sensing building change detection
Published in International Journal of Digital Earth, 2023
Xinyang Song, Zhen Hua, Jinjiang Li
FPT module. Here, we used three transformer types: self-transformer (ST), grounding-transformer (GT), and rendering transformer (RT). The three methods use different rules to fuse context information of different scales, as shown in Figure 6. The ST module only obtains object features that appear at the same time as the feature image. In essence, it is a modified non-local interaction (Wang et al. 2018), so that the output feature has the same ratio as the input X. Among them, a mixture of softmaxes (MoS) (Yang et al. 2017) is used as a normalization function, and then q and k are divided into n parts to calculate the similarity of each part score . The normalization function based on MoS can be expressed as where is the similarity score of the nth part, and is the nth aggregation weight. The GT module is a top-down non-local interaction, that maps the deep feature map onto the shallow feature map . Essentially, it enhances shallow features. The similarity is calculated using the euclidean distance instead of the dot product. Therefore, using as the similarity function is expressed as where and , is the ith feature position of , and represents the jth feature position of . More details on the functions and operations of the FPT module can be found in the Zhang and Shi (2020). The RT module works in a bottom-up manner, aiming to enhance the deep feature maps by rendering high-level semantic ”concepts” through low-level ”pixel” information. RT belongs to partial rendering, using high-level features to define Q, and low-level features to define K and V to highlight the rendering target. According to the rules of ST, GT, and RT, after calculating the corresponding features that integrate the context information of different scales, the feature maps are reordered according to size; finally, the final FPT features are generated through a convolutional layer.