This paper proposes a new reliability assessment method based on chip-level monitoring data. Traditional reliability assessment methods primarily rely on historical data statistics and generic failure rates of electronic components, facing challenges such as difficulty in obtaining reliability data, long testing periods, high costs, and insufficient accuracy due to assumptions. This study starts with chip-level monitoring and establishes a chip operation data acquisition system for secondary equipment in substations on a wide-area scale. By utilizing real-time operational data and combining it with failure data from equipment type tests, a failure threshold calculation method and scoring evaluation system are developed. This enables precise reliability assessment of secondary equipment in substations, providing a scientific basis for improving equipment reliability and optimizing operating conditions.

The new problem of multi-frequency oscillation and disturbance in new power system seriously threatens the safe operation of power grid. The existing multi-frequency phasor measurement technology can realize oscillation monitoring within 300Hz, centralized storage and monitoring at the master station, and off-line modeling and simulation method is used to evaluate the network risk. The large amount of multi-frequency phasor measurement data in the master station centralized storage and monitoring occupy a lot of resources, off-line risk assessment can not adapt to the real-time change of the power grid conditions. Based on the concept of edge computing, a multi-frequency oscillation collaborative online monitoring and advanced warning scheme is proposed to realize multi-level monitoring and advanced warning for bay-station-regional power grids. Key technologies such as grid bay monitoring, station area monitoring and advanced warning, master-substation coordination and wide-area coordination monitoring and advanced warning are expounded. The correctness and effectiveness of multi-frequency oscillation collaborative monitoring and advanced warning are verified by experiments. It can meet the requirement of re-al-time monitoring and risk advanced warning of multi-frequency oscillation in power grid.

This paper presents a comprehensive system architecture and device specifically designed for high-frequency signal monitoring for near HVDC converter station. The system is meticulously organized into two core modules: the master station and the substation. The master station is responsible for efficient data reception and panoramic monitoring analysis, and the substation includes a high-frequency signal synchronized measurement unit with a dedicated data processing unit. A novel spectrum correction method is proposed, which employs the mixed base FFT + ratio correction techniques and takes into account the filter compensation coefficient, leading to higher accuracy and precision. Through prolonged measurements on-site, it is confirmed that stable propagation of the 37th high-order harmonic exists at UHVDC near-field stations, and the dynamic trends of its amplitude and phase characteristics are revealed. In conjunction with power grid topology analysis, an in-depth exploration is conducted on the propagation characteristics of high-frequency signals and source localization strategies. The experimental results verify the prototype system's exceptional performance in terms of measurement stability and real-time capability, enabling wide-area synchronized monitoring of harmonics on-site and meeting the requirements of engineering applications.

This article addresses the problem of widespread harmonic frequency in the power grid caused by extensive integration of new energy sources into the new power system, which cannot be accurately measured by existing measurement and control equipment such as PMU. The article comprehensively applies transient quantities, phasors, effective values, and other algorithm models to design a wide-frequency measurement algorithm that supports cross-platform integration and deployment. The algorithm achieves unified measurement of power frequency, non-power frequency components, and oscillating power in the range of 0~2500Hz, enhances the detection capability of measuring devices for wide-frequency electrical quantities with different signal structures, and is deployed in multiple new energy power stations.

Composite visualization is a popular design strategy that represents complex datasets by integrating multiple visualizations in a meaningful and aesthetic layout, such as juxtaposition, overlay, and nesting. With this strategy, numerous novel designs have been proposed in visualization publications to accomplish various visual analytic tasks. However, there is a lack of understanding of design patterns of composite visualization, thus failing to provide holistic design space and concrete examples for practical use. In this article, we opted to revisit the composite visualizations in IEEE VIS publications and answered what and how visualizations of different types are composed together. To achieve this, we first constructed a corpus of composite visualizations from the publications and analyzed common practices, such as the pattern distributions and co-occurrence of visualization types. From the analysis, we obtained insights into different design patterns on the utilities and their potential pros and cons. Furthermore, we discussed usage scenarios of our taxonomy and corpus and how future research on visualization composition can be conducted on the basis of this study.

Dialogue system is designed to converse with humans in a natural way. As an essential part of dialogue system, dialogue generation aims to generate proper response given historical context. Recently, sequence-to-sequence (seq2seq) based models have achieved great success but suffer from ungrammatical problems. In this paper, we propose a Syntax-aware Dialogue Generation (SynDG) model that incorporates syntactic information to generate grammatical responses with an encoder-decoder framework. Specifically, we first construct a syntax-graph with a dependency parser on the dialogue corpus. Then, we employ three graph embedding algorithms to learn syntactic word representations as the input of seq2seq framework. Furthermore, we devise training strategies to predict syntactic structure of the sentence for sufficient syntax understanding. Our empirical study on two multi-turn dialogue datasets demonstrates the effectiveness of SynDG in generating natural and grammatical responses.

Document grounded conversation (DGC) aims to generate informative responses when talking about a document. It is normally formulated as a sequence-to-sequence (Seq2seq) learning problem, which directly maps source sequences, i.e., the context and background documents, to the target sequence, i.e., the response. These responses are normally used as the final output without further polishing, which may suffer from the global information loss owing to the auto-regression paradigm. To tackle this problem, some researches designed two-pass generation to improve the quality of responses. However, these approaches lack the capability of distinguishing inappropriate words in the first pass, which may maintain the erroneous words while rewrite the correct ones. In this paper, we design a scheduled error correction network (SECN) with multiple generation passes to explicitly locate and rewrite the erroneous words in previous passes. Specifically, a discriminator is employed to distinguish erroneous words which are further revised by a refiner. Moreover, we also apply curriculum learning with reasonable learning schedule to train our model from easy to hard conversations, where the complexity is measured by the number of decoding passes. We conduct comprehensive experiments on a public document grounded conversation dataset, Wizard-of-Wikipedia, and the results demonstrate significant promotions over several strong benchmarks.

Formulating dialogue policy as a reinforcement learning (RL) task enables a dialogue system to act optimally by interacting with humans. However, typical RL-based methods normally suffer from challenges such as sparse and delayed reward problems. Besides, with user goal unavailable in real scenarios, the reward estimator is unable to generate reward reflecting action validity and task completion. Those issues may slow down and degrade the policy learning significantly. In this paper, we present a novel scheduled knowledge distillation framework for dialogue policy learning, which trains a compact student reward estimator by distilling the prior knowledge of user goals from a large teacher model. To further improve the stability of dialogue policy learning, we propose to leverage self-paced learning to arrange meaningful training order for the student reward estimator. Comprehensive experiments on Microsoft Dialogue Challenge and MultiWOZ datasets indicate that our approach significantly accelerates the learning speed, and the task-completion success rate can be improved from 0.47%similar to 9.01% compared with several strong baselines.

高比例新能源和高比例电力电子设备的接入,产生了大量间谐波信号,并引发一系列宽频域振荡事件,严重影响了电网的运行安全,现有测量技术局限于工频信号测量,无法实时监测宽频振荡等间谐波信号.文中以《电力系统宽频测量装置技术规范》发布为契机,介绍了宽频测量技术的提出背景、装置功能定位及相关术语定义,从宽频信号的采样频率、数据测量、宽频振荡监测、数据传输以及数据录波等方面讨论了宽频测量装置的功能和性能,论述了装置工程应用的方案、应用场景和应用展望,总结了标准制定中存在的问题及下一步研究方向,为宽频测量装置的工程应用提供指导和参考.

In this work, thermo-mechanically treated 42CrMo steel was subjected to cryogenic treatment conducted by means of orthogonal design method, followed by low-temperature tempering to investigate the effect of different parameters of cryogenic treatment on wear resistance of 42CrMo steel and to optimize parameters of cryogenic treatment for improving wear resistance. The results of hardness test and wear test show that cryogenic treatment significantly improves wear resistance with marginal changes in coefficient of friction and hardness. Specifically, cryogenic temperature has the largest impact on wear resistance of 42CrMo steel, holding time has medium impact, and the parameter of treatment cycles has the least impact. The optimum parameters of cryogenic treatment are −196°C for 12 hours with one cycle for improving wear resistance. The results of scanning electron microscopy (SEM) and X-ray diffractometry (XRD) analysis indicate that marginal changes in hardness and coefficient of friction may be owing to little amount of transformation of retained austenite, and the significant influence of cryogenic treatment on improving wear resistance of 42CrMo steel can be mainly attributed to segregation of carbon atoms promoted by cryogenic treatment resulting in more precipitation of carbides in subsequent tempering.

Path planning plays a vitally important role for agents in adversarial strategy games, and even directly determines the final success or failure at times. However, most previous studies of path planning focus on minimizing path length or cost, ignoring the growing complexity resulting from dynamic changing situations. In this paper, we propose a multi-factor situational coupling D* lite (MFSCD* lite) to address the issues above. First, this paper models important dynamic situational factors and integrates them into MFSCD* lite, allowing the planning path to dynamically fit with the changing of situation. Secondly, the buffer and extended buffer of factor-values are established to reduce the repeated calculation of situational factors in MFSCD* lite. And the update strategy is optimized to improve the efficiency of path planning. To validate the effectiveness of our method, we conduct comprehensive experiments on wargame, a typical real-time agent adversarial game. The experimental results show that MFSCD* lite can meet the needs of high-quality and fast path planning, which is better than ordinary path planning algorithms. Further, it has achieved excellent performance in major wargame competitions.

In War-Game, the opponent’s location is the key information for decision-making, but it is hard to be completely known because of "the fog of war". Traditional location prediction methods model sequential behavior of a single target and predict the next location based on its recent locations. This restricted form of modeling benchmark limits the location prediction accuracy in War-game since explicitly omitting the fact that the movement of a target unit is affected by the situation and other units. In this paper, we propose a spatial-temporal heterogeneous graph neural network to encode semantic relation between combat units and integrate other units’ information into opponent's location prediction. The historical situation is modeled as a spatial-temporal graph, of which heterogeneous edges are used to describe different semantic relation of combat units. Furthermore, we apply spatial graph convolution block and temporal gated recurrent units for situation representation learning. To validate the effectiveness of our method, we construct two benchmark datasets with different scenarios and conduct comprehensive experiments on them. Experiments show that our proposed method extracts the correlated information of location prediction from other combat units and achieve 5.13% and 18.70% improvements over the strongest baselines.

针对从"人在回路"兵棋推演的复盘数据中提取推演者战术经验高价值知识的问题,提出一种基于深度神经网络从复盘数据中学习战术机动策略模型的方法.将战术机动策略建模为在当前态势特征影响下对目标候选位置进行优选的分类问题:梳理总结影响推演者决策的关键认知因素,定义了由机动范围和观察范围等7个属性构成的基础态势特征,建立了带有正负样本标注的态势特征数据集;设计了基于卷积神经网络的分类器,以分类概率实现了单个棋子战术机动终点位置的预测.实验结果表明:该模型的预测准确率可达到78.96％,相比其他模型提高至少4.59％.

Popularity prediction of online content over social media platforms is a valuable and challenging issue, the core of which lies in how to capture predictive factors from available data. However, existing studies either treat each cascade independently, which neglects the correlation among different cascades, or lack a comprehensive consideration of user behavioral proximity and preference with respect to different messages. Motivated by the above observation, this article proposes a graph neural network-based framework named HeDAN (heterogeneous diffusion attention network), which comprehensively considers various factors affecting information diffusion to provide more accurate prediction results. Specifically, we first construct a heterogeneous diffusion graph with two types of nodes (user and message) and three types of relations (friendship, interaction, and interest). Among them, friendship reflects the strength of social relationships between users, interaction reflects the behavioral proximity between users, and interest reflects user preference for information. Next, a graph neural network model with a hierarchical attention mechanism is proposed to learn from these relations. Specifically, at the node level, we utilize the graph attention network to learn the subgraph structure and generate the representations of users and messages under each specific relationship. At the semantic level, we distinguish the importance of different nodes in different relations via the multi-head self-attention mechanism and fuse them into the final prediction representation. Extensive experimental results on three real diffusion datasets show the superior performance of HeDAN over the state-of-the-art baselines.

Writing formulas on the spreadsheet grid is arguably the most widely practiced form of programming. Still, studies highlight the difficulties experienced by end-user programmers when learning and using traditional formulas, especially for slightly complex tasks. The purpose of GridBook is to ease these difficulties by supporting formulas expressed in natural language within the grid; it is the first system to do so. GridBook builds on a parser utilizing deep learning to understand analysis intents from the natural language input within a spreadsheet cell. GridBook also leverages the spatial context between cells to infer the analysis parameters underspecified in the natural language input. Natural language enables users to analyze data easily and flexibly, to build queries on the results of previous analyses, and to view results intelligibly within the grid-thus taking spreadsheets one step closer to computational notebooks. We evaluated GridBook via two comparative lab studies, with 20 data analysts new only to GridBook. In our studies, there were no significant differences, in terms of time and cognitive load, in participants' data analysis using GridBook and spreadsheets; however, data analysis with GridBook was significantly faster than with computational notebooks. Our study uncovers insights into the application of natural language as a special purpose programming language for end-user programming in spreadsheets.

Dialogue state tracking (DST), as an essential component of task-oriented dialogue systems, refers to keeping track of the user's intentions as a conversation progresses. Typical methods formulate it as a classification task with fixed pre-defined slot-value pairs, or generate slot-value candidates given the dialogue history. Most of them have limitations on considering interactions of slots with utterance sentences and other slots progressively. To tackle this problem, we propose a Dialogue State Tracker with Hierarchical Temporal Slot Interactions (DST-HTSI) to capture slot-related semantic information from utterance sentences and slots. It firstly captures interactive information among slots within a turn and across turns by applying hierarchical slot interactions. Then a temporal slot interaction module is employed to establish slot dependencies along the time. Finally, a GRU is applied as the decoder to generate values for each slot correspondingly. Furthermore, we also leverage pre-trained language models as the backbone of our model. Experiments show that DST-HTSI outperforms previous state-of-the-art on MultiWOZ 2.2 and WOZ 2.0, and achieves competitive results on MultiWOZ 2.1.

The large-scale integration of new energy sources has brought a series of new types of oscillations to the power grid, and the oscillations gradually develop to high frequencies. The existing WAMS-based oscillation monitoring technology can only monitor sub-synchronous oscillation below 50Hz, and cannot cope with wide-frequency oscillations above 50Hz. The existing monitoring methods can only passively respond to the oscillation after the oscillation occurs, and lack active early warning and intervention methods. Based on wide-frequency measurement technology, this paper proposes an engineering method for wide-frequency oscillation risk assessment and early warning of oscillation. Firstly, it discusses the functions and characteristics of wide-frequency measurement technology and devices, and proposes an overall scheme for oscillation risk assessment and early warning based on wide-frequency measurement data. Then, the method of oscillation risk assessment and early warning is discussed in detail from two aspects, which are the statistical analysis of historical data and real-time monitoring data, it can provide guidance and reference for the early warning and analysis of power grid wide-frequency oscillation in the future, and promote oscillation monitoring from passive response to active prevention and intervention.

Distributional text clustering delivers semantically informative representations and captures the relevance between each word and semantic clustering centroids. We extend the neural text clustering approach to text classification tasks by inducing cluster centers via a latent variable model and interacting with distributional word embeddings, to enrich the representation of tokens and measure the relatedness between tokens and each learnable cluster centroid. The proposed method jointly learns word clustering centroids and clustering-token alignments, achieving the state of the art results on multiple benchmark datasets and proving that the proposed cluster-token alignment mechanism is indeed favorable to text classification. Notably, our qualitative analysis has conspicuously illustrated that text representations learned by the proposed model are in accord well with our intuition.

A large-scale integration of renewable energy of grid has introduced a large number of power electronic equipment, which introduced a lot of inter-harmonics and harmonic signals into the grid, and has shown the trend of power electronics. The existing measurement technology only focuses on 50Hz power frequency signal, and cannot meet the real-time measurement requirements of inter-harmonic and harmonic. Wide-frequency measurement device can realize the unified monitoring of the fundamental wave, inter-harmonic and harmonic of the power grid, but a large amount of data cannot be transmitted to the master station in real time, and it needs to be processed and analyzed locally. Firstly, this article analyzes the type of wide-frequency measurement data, the data filtering mechanism. Secondly, it discusses the preprocessing and analysis of the wide-frequency measurement data and oscillation information on substation level, focusing on the substation-level steady-state real-time data preprocessing analysis and alarm event analysis. Thirdly, this paper discusses the realization of the preprocessing analysis report and alarm event analysis report, and then discusses the transmission and engineering application of the analysis reports. It can provide guidance and reference for the wide-frequency monitoring and engineering application of the power electronics dominated power system in future.

Generative Adversarial Networks (GANs) have achieved great success in image synthesis, but have proven to be difficult to generate natural language. Challenges arise from the uninformative learning signals passed from the discriminator. In other words, the poor learning signals limit the learning capacity for generating languages with rich structures and semantics. In this paper, we propose to adopt the counter-contrastive learning (CCL) method to support the generator’s training in language GANs. In contrast to standard GANs that adopt a simple binary classifier to discriminate whether a sample is real or fake, we employ a counter-contrastive learning signal that advances the training of language synthesizers by (1) pulling the language representations of generated and real samples together and (2) pushing apart representations of real samples to compete with the discriminator and thus prevent the discriminator from being overtrained. We evaluate our method on both synthetic and real benchmarks and yield competitive performance compared to previous GANs for adversarial sequence generation.