The smart grid aims to provide economically efficient and sustainable power to consumers with high quality and security. However, the increasing integration of distributed renewable energy sources presents challenges for smart grid protection and control systems. To enhance smart grid operations, this article introduces White Rabbit (WR) as a more precise and accurate time synchronization technique. To explore White Rabbit’s potential applications in the smart grid, this study first explains existing time synchronization techniques and their limitations, followed by an analysis of the role and importance of time synchronization in smart grids. A comprehensive survey is then conducted, covering the theories, principles, implementations, performances, and existing application cases of White Rabbit. The findings suggest that White Rabbit is a promising technique for smart grid deployment. However, several challenges remain on the path to large-scale implementation. These challenges are analyzed in detail, highlighting key areas for future research.
4 January 2018 in IEEE Transactions on Power Delivery
An oscillation blocking and re-opening scheme for distance protection in multi-source oscillation scenes is proposed in this paper. First, the expression of measured impedance in multi-source oscillation scenes is derived and the variation characteristic of the trajectory of measured impedance is analyzed. Then, according to the correlation between the measured impedance and the position of desynchronizing center, the desynchronizing center position factor (DCPF) is defined, and protection blocking and re-opening criterion based on DCPF is constructed. RTDS (Real Time Digital Simulator) test results demonstrate that, the proposed scheme can effectively block the protection during oscillation even when fault occurs at the same time as the desynchronizing center appears, and re-open the protection quickly when fault occurs.
Large-area boron nanowire (BNW) films were fabricated on the Si(111) substrate by chemical vapor deposition (CVD). The average diameter of the BNWs is about 20 nm, with lengths of 5–10 μm. Then, graphene-capped boron nanowires (GC-BNWs) were obtained by microwave plasma chemical vapor deposition (MPCVD). Characterization by scanning electron microscopy indicates that few-layer graphene covers the surface of the boron nanowires. Field emission measurements of the BNWs and GC-BNW films show that the GC-BNW films have a lower turn-on electric field than the BNW films.
The effects of different kinds of communication channels on the key indexes such as bit error rate and communication rate of quantum communication system, represented by quantum secure communication system, are systematically studied. For the actual channel, the transformation of quantum states caused by physical effects such as loss, dispersion and polarization mode dispersion in optical fiber channel is analyzed. With building in the spatial channel noise experimental data under different weather conditions, the mathematical model is established, and the software is designed and implemented.
Various robustness evaluation methodologies from different perspectives have been proposed for different natural language processing (NLP) tasks. These methods have often focused on either universal or task-specific generalization capabilities. In this work, we propose a multilingual robustness evaluation platform for NLP tasks (TextFlint) that incorporates universal text transformation, task-specific transformation, adversarial attack, subpopulation, and their combinations to provide comprehensive robustness analysis. TextFlint enables practitioners to automatically evaluate their models from all aspects or to customize their evaluations as desired with just a few lines of code. To guarantee user acceptability, all the text transformations are linguistically based, and we provide a human evaluation for each one. TextFlint generates complete analytical reports as well as targeted augmented data to address the shortcomings of the model's robustness. To validate TextFlint's utility, we performed large-scale empirical evaluations (over 67,000 evaluations) on state-of-the-art deep learning models, classic supervised methods, and real-world systems. Almost all models showed significant performance degradation, including a decline of more than 50% of BERT's prediction accuracy on tasks such as aspect-level sentiment classification, named entity recognition, and natural language inference. Therefore, we call for the robustness to be included in the model evaluation, so as to promote the healthy development of NLP technology.