Method of Voltage Sag Causes Based on Bidirectional LSTM and Attention Mechanism

Hong Wang,Linhai Qi,Yongshuo Ma,Jiehui Jia,Zhicong Zheng 대한전기학회 2020 Journal of Electrical Engineering & Technology Vol.15 No.3

The voltage sags’ cause recognition is the basis for formulating governance plans and clarifying liabilities for accidents. For voltage sag cause recognition methods based on physical characteristics, new challenges are presented in terms of accuracy, adaptability and algorithm efciency. Deep learning is a method based on characterizing and learning data. The efcient mechanism of autonomous feature learning can efectively overcome the problems of information loss and generalization ability based on existing physical property methods. The long short-term memory network (LSTM) has the characteristics of memory and can better learn the data characteristics with time series characteristics. Bidirectional LSTM can consider historical information and future information compared with standard LSTM, and has more processing for time series data. While using the attention mechanism can highlight the key infuencing factors in the time series and improve the recognition accuracy of the model. For the transient sag time series data, this paper proposes a multi-layer structure based on bidirectional LSTM and attention mechanism to classifcation recognition. The experiment uses simulation data and measured data to prove that the model has good recognition ability and good anti-noise performance in the recognition of voltage sag causes, and can be reliably applied in practical engineering.

Numerical study on CMT boron replenishment strategy for an AP1000 nuclear power unit

Hong Wang,Miao Zhang,Jialong Li,Junpeng Wang 한국원자력학회 2022 Nuclear Engineering and Technology Vol.54 No.6

The passive safety system is adopted in an AP1000 nuclear power unit to improve the operation safety ofthe whole unit. However, due to boron diffusion and periodic sampling, boron dilution occurs in the coremakeup tank. The boron replenishment process in the core makeup tank is essential and becomesparticularly important. Based on the validated models, this article numerically investigates the influenceof the replenishment flow rate and the position on the boron distribution in the core makeup tank. Thethermal fatigue phenomenon of the “T” connection caused by replenishment is analyzed. Finally, thereplenishment strategy is proposed to benefit both boron mixing in the core makeup tank and eliminating the thermal fatigue of the “T” connection

Decision making with distance and cosine similarity measures for intuitionistic hesitant fuzzy sets

Hong Wang,Dandan Li 원광대학교 기초자연과학연구소 2017 ANNALS OF FUZZY MATHEMATICS AND INFORMATICS Vol.13 No.6

The definition of intuitionistic hesitant fuzzy sets(IHFSs) is developed based on intuitionistic fuzzy sets(IFSs) and hesitant fuzzy sets(HFSs) and it allows the membership of an element to be a set of several possible intuitionistic fuzzy values. The intuitionistic hesitant fuzzy sets(IHFSs) is a new and flexible tool in representing hesitant information in decision making. Distance and cosine similarity measures have been applied widely in many research domains and practical fields. In this paper, we firstly proposed some distance and similarity measures for IHFSs based on Hamming distance, Euclidean distance and generalized distance, especially, a new cosine similarity measure for IHFSs is proposed and the corresponding cosine distance measures are given. It is shown that all three parameters (membership degree, non membership degree and degree of hesitation) describing intuitionistic hesitant fuzzy sets should be taken into account when calculating those distance and cosine similarity measures. In the end, a new cosine similarity measure and distance measures are applied to multiple attribute decision-making.

AHP 기법 및 KANO 모델을 활용한 도시 수변 공공공간 만족도 평가 - 부산시 APEC 나루공원을 중심으로 -

Hong Wang,조정형 사단법인 한국융합기술연구학회 2023 아시아태평양융합연구교류논문지 Vol.9 No.6

Urban parks are open spaces in the city that harmonize with nature, providing places for citizens to take walks, exercise, and leisure, and playing an essential role in maintaining and preserving the urban ecosystem. Therefore, in the design process, the goal pursued by designers is to increase public space satisfaction centered on users. This study evaluated the park from the perspective of user satisfaction, focusing on APEC Naru Park, a typical waterfront public space. In this paper, a quantitative method for urban parks was adopted by introducing a Kano model combined with POE and AHP analysis methods, and an improved evaluation index system was established to clarify the factors that APEC Naru Park should improve and the urgency of improvement. Through this study, it is possible to provide a certain basis necessary to improve the quality of urban public spaces, including APEC Naru Park in Busan. It is expected that this can be used as a reference for improving public spaces in other cities. Through this, it will be able to contribute to improving the quality of urban space and creating a better urban environment that can improve the quality of life of citizens. 도시공원은 도시에서 자연과 조화를 이루는 개방된 공간으로 시민들이 산책, 운동, 레저 등을 취할 수 있는 장소를 제공하며, 도시 생태계를 유지하고 보존하는 데 필수적인 역할을 한다. 따라서 디자인 과정에서는 사용자 중심으로 공공공간 만족도를 높이는 것이 설계자들이 추구하는 목표이다. 본 연구는 전형적인 수변 공공공간인 APEC 나루공원을 중심으로 하여 사용자 만족도의 관점에서 공원을 평가하였다. 본 논문에서는 POE(Post Occupancy Evaluation) 및 AHP(The Analytic Hierarchy Process) 계층분석법과 결합된 Kano 모델을 도입하여 도시공원에 대한 정량적 방법을 채택하였으며, 개선된 평가지표 시스템을 구축하여 APEC 나루공원이 개선해야 할 요소와 개선의 시급성을 명확히 한다. 본 연구를 통해 부산시 도시공원의 도시 공공공간 품질을 향상시키는데 필요한 일정한 기반을 제공할 수 있다. 이는 다른 도시의 공공공간 개선을 위한 참고 자료로 활용될 수 있을 것으로 기대된다. 이후 도시공간의 품질을 개선하고 시민들의 삶의 질을 높일 수 있는 더 나은 도시환경을 조성하는데 기여할 수 있을 것이다.

A Novel Globally Adaptive Load-Balanced Routing Algorithm for Torus Interconnection Networks

Hong Wang,Du Xu,Lemin Li 한국전자통신연구원 2007 ETRI Journal Vol.29 No.3

A globally adaptive load-balanced routing algorithm for torus interconnection networks is proposed. Unlike previously published algorithms, this algorithm employs a new scheme based on collision detection to handle deadlock, and has higher routing adaptability than previous algorithms. Simulation results show that our algorithm outperforms previous algorithms by 16% on benign traffic patterns, and by 10% to 21% on adversarial traffic patterns.

A Selective Review on Random Survival Forests for High Dimensional Data

Hong Wang,Gang Li 계명대학교 자연과학연구소 2017 Quantitative Bio-Science Vol.36 No.2

Over the past decades, there has been considerable interest in applying statistical machine learning methods in survival analysis. Ensemble based approaches, especially random survival forests, have been developed in a variety of contexts due to their high precision and non-parametric nature. This article aims to provide a timely review on recent developments and applications of random survival forests for time-to-event data with high dimensional covariates. This selective review begins with an introduction to the random survival forest framework, followed by a survey of recent developments on splitting criteria, variable selection, and other advanced topics of random survival forests for time-to-event data in high dimensional settings. We also discuss potential research directions for future research.

DISJOINT CYCLES WITH PRESCRIBED LENGTHS AND INDEPENDENT EDGES IN GRAPHS

Hong Wang 대한수학회 2014 대한수학회지 Vol.51 No.5

We conjecture that if k ≥ 2 is an integer and G is a graph of order n with minimum degree at least (n + 2k)/2, then for any k independent edges e1 ,...,ek in G and for any integer partition n = n1 + ··· + nk with ni ≥ 4 (1 ≤ i ≤ k), G has k disjoint cycles C1 ,...,Ck of orders n1 ,...,n k , respectively, such that C i passes through e i for all 1 ≤ i ≤ k. We show that this conjecture is true for the case k = 2. The minimum degree condition is sharp in general.

Microphysical Structures of an Extreme Rainfall Event Over the Coastal Metropolitan City of Guangzhou, China: Observation Analysis with Polarimetric Radar

Hong Wang,Jin-Fang Yin,Naigeng Wu,Weiyu Ding 한국기상학회 2023 Asia-Pacific Journal of Atmospheric Sciences Vol.59 No.1

A record-breaking nocturnal rainfall event (543 mm in 16-h) under weak synoptic forcing occurred in the metropolitan city of Guangzhou, China, during 6–7 May 2017. The evolution and microphysical structures of this torrential rainfall event are investigated using S-band polarimetric radar datasets. The torrential rainfall concentrated in two cores: one over Huadu District (HD) in which the storms were initiated between urban areas and mountains at mid-night, and the other over Huangpu and Zengcheng District (ZC) which was characterized by locally triggered storms merging with the storms from HD. The two heavy precipitation regions show some similarities, including strong reflectivity factor for horizontal polarizations (ZH) magnitude, low centroid cumulonimbus structures, and column shape of differential reflectivity (ZDR). But obvious differences can also be viewed between them. Compared to HD, ZC has higher precipitation intensity, longer precipitation duration, and larger accumulated rainfall. Besides, ZC also has a relatively lower ZDR value of ~ 0.2 dB and a higher specific differential phase (KDP) of approximately ~ 0.35° km− 1, which indicates the larger population of medium-sized rain droplet and higher water content in ZC. The radar-retrieved drop size distributions (DSDs) (i.e., mass-weighted diameter, logarithmic normalized intercept, and liquid water content) show that small size particles and high particle number concentration are more obvious in the storm over ZC. Combined with the retrieved DSDs, the merger process brings more medium-sized raindrops to ZC, and increases the possibility of raindrop growth via the accretion of cloud water by rain, which leads to enhancement of precipitation. In addition, strong KDP may be a good indicator of intensity for extreme precipitation.

A Network Intrusion Security Detection Method Using BiLSTM-CNN in Big Data Environment

Hong Wang 한국정보처리학회 2023 Journal of information processing systems Vol.19 No.5

The conventional methods of network intrusion detection system (NIDS) cannot measure the trend of intrusiondetectiontargets effectively, which lead to low detection accuracy. In this study, a NIDS method which basedon a deep neural network in a big-data environment is proposed. Firstly, the entire framework of the NIDSmodel is constructed in two stages. Feature reduction and anomaly probability output are used at the core ofthe two stages. Subsequently, a convolutional neural network, which encompasses a down sampling layer anda characteristic extractor consist of a convolution layer, the correlation of inputs is realized by introducingbidirectional long short-term memory. Finally, after the convolution layer, a pooling layer is added to samplethe required features according to different sampling rules, which promotes the overall performance of theNIDS model. The proposed NIDS method and three other methods are compared, and it is broken down underthe conditions of the two databases through simulation experiments. The results demonstrate that the proposedmodel is superior to the other three methods of NIDS in two databases, in terms of precision, accuracy, F1-score, and recall, which are 91.64%, 93.35%, 92.25%, and 91.87%, respectively. The proposed algorithm issignificant for improving the accuracy of NIDS.

Bipacking a bipartite graph with girth at least 12

Hong Wang 대한수학회 2019 대한수학회지 Vol.56 No.1

Let $G$ be a bipartite graph with $(X,Y)$ as its bipartition. Let $B$ be a complete bipartite graph with a bipartition $(V_1,V_2)$ such that $X\subseteq V_1$ and $Y\subseteq V_2$. A {\em bi-packing} of $G$ in $B$ is an injection $\sigma$: $V(G) \rightarrow V(B)$ such that $\sigma (X)\subseteq V_1$, $\sigma (Y)\subseteq V_2$ and $E(G)\cap E(\sigma (G))=\emptyset$. In this paper, we show that if $G$ is a bipartite graph of order $n$ with girth at least 12, then there is a complete bipartite graph $B$ of order $n+1$ such that there is a bi-packing of $G$ in $B$. We conjecture that the same conclusion holds if the girth of $G$ is at least 8.