Journal of University of Chinese Academy of Sciences

Continuity of truncated Hardy-Littlewood maximal function

WANG Yidong, WU Jia, YAN Dunyan

2024, 41 (6): 721-727. DOI: 10.7523/j.ucas.2023.045

This paper focuses on the continuity of the truncated Hardy-Littlewood maximal function. We first show that the truncated Hardy-Littlewood maximal function is lower semi-continuous. Then by investigating the behavior of the truncated Hardy-Littlewood maximal function when the truncated parameter γ changes, we obtain an equivalent condition of the continuity of the truncated Hardy-Littlewood maximal function.

Interval estimation for the process capability index under the inverse Rayleigh and log-logistic distributions

MU Weiyan, JIA Xiaofang, XIONG Shifeng

2024, 41 (6): 728-735. DOI: 10.7523/j.ucas.2023.001

Abstract ( 124 ) PDF (0KB) ( 0 )

In this paper， we discuss the statistical inferential problem on an important process capability index C_Npk under the inverse Rayleigh and log-logistic distributions. We construct the generalized pivotal quantities for C_Npk, and subsequently the generalized confidence intervals proposed. Numerical results show that the proposed interval estimation performs better than the bootstrap method in terms of coverage rate and interval width.

Numerical study of natural convection heat transfer from neck folds of the frill-neck lizard

JIA Chongxi, WANG Hao, LIU Jie, LU Wenqiang

2024, 41 (6): 736-745. DOI: 10.7523/j.ucas.2022.078

Abstract ( 302 ) PDF (0KB) ( 0 )

In this paper, the physical modeling of frill-neck lizard, a reptile with special heat dissipation structure, was carried out, and the natural convection heat transfer properties of it were numerically studied. It was found that there was only one maximum heat transfer coefficient in the range of 45°-85°, which was obtained at around 65° for the simulation of different field angle adjustments of frill-neck lizard neck fold. At the same time, by changing the size of frill-neck lizard neck fold, it was also observed that there was a unique maximum convection heat transfer coefficient in the range of 6.3≤ L₀/d ≤6.6. In addition, the field angle and area corresponding to the maximum heat transfer rate were close to the natural state of Australian subspecies. Therefore, it can be inferred that the evolution direction of frill-neck lizard neck fold may be beneficial to improving the natural convection heat transfer rate. Hence, the natural convection between frill-neck lizard and external environment may be considered as an important reason affecting the evolution direction of Chlamydosaurus kingii neck fold.

Experimental study of MHD effect of phase change heat transfer in metals under the influence of a strong magnetic field

CAI Zhiyang, MENG Xu, ZHANG Dengke, WU Xi, WANG Zenghui

2024, 41 (6): 746-754. DOI: 10.7523/j.ucas.2023.021

Abstract ( 267 ) PDF (0KB) ( 0 )

As a highly efficient heat transport medium, the study of the melting and heat transfer characteristics of metallic fluids in phase change processes under magnetic fields is of great importance for industrial processes such as fusion reactors, electromagnetic metallurgy, and additive manufacturing. In this paper, the melting process of metallic gallium under a strong magnetic field was studied by building a comprehensive experimental system for heat transfer through phase change of metal, and the heat transfer characteristics of metallic gallium melting under the action of a magnetic field were obtained. The dynamic average distance of the heated wall from the phase interface during melting instead of the fixed characteristic length was used to study the variation of the relative strength of convective heat transfer and thermal conductivity with Fourier number (Fo) during melting. The results show that: under a small Hartmann number (Ha), the melting has a melting-promoting effect at the early stage and is inhibited at the later stage; under a large Hartmann number the magnetic field has an inhibiting effect on the convection during the melting of gallium metal, and the melting process shows a laminar and uniform advance. The magnetic field reduces the height of the dominant zone of thermal conductivity at the bottom of the cavity during the melting process and suppresses temperature fluctuations during the melting process, resulting in a uniform temperature distribution during the melting process.

Comparison study on classification accuracy of 11 common water indices based on Landsat 8 OLI images

LI Longjie, YANG Yonghui

2024, 41 (6): 755-765. DOI: 10.7523/j.ucas.2023.088

Abstract ( 378 ) PDF (0KB) ( 0 )

Water index is one of the most effective methods to extract water bodies from remote sensing images. There are many kinds of water index, each with its own characteristics. It is, therefore, necessary to select the index with best classification accuracy. Taking Shijiazhuang City as the research area, 11 common water indices were used to extract water bodies from Landsat 8 OLI images. The accuracy of the water index extraction results is validated by using the visual interpretation (VI) result as the standard classification map from Sentinel-2 MSI based on the area test method in combination with transition matrix and sampling test method. Results show little difference in the extraction of large water bodies among different water indices. Small ponds and rivers can better check the extraction ability of water index. It is proved that Water Index 2019 (WI2019) has the best water classification. WI2019 is then used to find out the recent expansion of water bodies after the start of South-to-North Water Diversion Project for water transfer. It was found that the area of surface water body in Shijiazhuang excluding large reservoirs increased significantly, from 42 km² in 2014 to 62 km² in 2020, an increase of 20 km². In view of the canal seepage control treatment at the bottom of most newly added water bodies, with poor groundwater recharge function, and more ineffective evaporation, it is recommended to properly control the scale of water bodies in order to effectively reduce the waste of water transferred from outside.

Measurement and influencing factors of open economy in border cities： taking Hunchun City, China as an example

LIAO Maowei, ZHANG Pingyu, LI Yuxin

2024, 41 (6): 766-775. DOI: 10.7523/j.ucas.2023.026

Abstract ( 195 ) PDF (0KB) ( 0 )

Based on the concept connotation of open economy, this paper builds the evaluation index system for open economy in border cities from opening to the outside world and opening to the interior. Taking Hunchun City, China as an example, the entropy method, coupling coordination model and obstacle degree model were used to explore the development process, characteristics and influencing factors of open economy in Hunchun from 2001 to 2020. The results are as follows: 1) The economic openness of Hunchun showed a trend of fluctuating upward, and the opening process has experienced three stages: slow development (2001-2008), rapid development (2009-2014) and stable development (2015-2020). Opening to the outside world and the interior have gradually transitioned from barely coordinated to well-coordinated. 2) During the study period, the obstacle degree of opening to the outside world showed a trend of fluctuating upward, while the obstacle degree of opening to the interior showed a trend of fluctuating downward. Opening to the outside world is the critical factor that restricts the opening development of Hunchun at its current stage. The main obstacle factors include degree of dependence on foreign trade, the number of newly approved foreign-funded enterprises, actual utilization of foreign capital and investment promotion.

Lightweight network for fast ship detection in SAR images

ZHOU Wenxue, ZHANG Huachun

2024, 41 (6): 776-785. DOI: 10.7523/j.ucas.2023.017

Abstract ( 358 ) PDF (0KB) ( 0 )

In the field of SAR image ship detection based on deep learning, traditional models are usually complex in structure and require a large amount of calculation, making them unsuitable for low computing power platforms and real-time detection. And convolutional neural networks that rely on preset anchor boxes will lead to a lot of computational redundancy due to the difficulty of setting a reasonable anchor box. To solve these problems, an end-to-end lightweight convolutional neural network based on anchor-free design is proposed, and a lightweight channel attention module (EESE) is designed and applied to the detection head (ED-head), to resolve the conflict between classification and localization tasks. In addition, an optimized EIOU loss function is proposed, which enables the model to effectively improve the network performance without increasing the inference time. The proposed method is tested on the SSDD dataset, and the experimental results show that compared to YOLOX-nano, AP₅₀ and AP are increased by 2.1 and 7.4 percentage points, respectively, with the CPU latency being only 5.33 ms, much less than 13.13 ms of YOLOX-nano. The proposed method achieves a balance between accuracy and efficiency.

The improved SLM algorithm used in hybrid beamforming architecture

XIAO Disheng, HU Shicheng, QIAN Hua, KANG Kai, LI Mingqi

2024, 41 (6): 786-793. DOI: 10.7523/j.ucas.2023.015

Abstract ( 179 ) PDF (0KB) ( 0 )

In the massive multi-input multi-output system, the peak-to-average power ratio (PAPR) is one of the factors which greatly affect the performance of the transmitter. The existing PAPR reduction methods are based on the fully-digital architecture, which can not effectively reduce the PAPR of signal at the transmitting antennas in the hybrid beamforming architecture. To address this problem, an improved selective mapping (SLM) method is proposed, which adopts independent phase rotations to the initial input signal and then calculates the PAPR at the transmitting antennas and finally transmits the sequence with minimum PAPR. Besides, the upper bound and lower bound of PAPR at the transmitting antennas are analyzed. Theoretical analysis and simulation results suggest that the proposed improved SLM can effectively reduce the PAPR at the transmitting antennas in the hybrid beamforming architectures.

Water image extraction algorithm based on improved Gaussian mixture model and graph cut model

BAO Linan, LYU Xiaolei

2024, 41 (6): 794-802. DOI: 10.7523/j.ucas.2023.028

Abstract ( 196 ) PDF (0KB) ( 0 )

Synthetic aperture radar (SAR) has the characteristics of all-day and all-weather imaging, wide observation range, and short mapping period, which make it highly advantageous in water extraction. However, existing algorithms for lake extraction are easily affected by the surrounding environment of lakes and noise interference, resulting in low operational efficiency. Therefore, this paper proposes a detection method that combines an improved Gaussian mixture model (GMM) with graph cut model (GCM). First, the two-level Otsu threshold method is used to obtain the initial segmentation map of the lake, and the calculated parameter set is used as the initial parameter of the GMM. The expectation maximum algorithm (EM) is employed to obtain the optimal parameters of the GMM iteratively. The experimental results demonstrate that the more accurate the initial parameters, the clearer the outline of the water body. The introduction of the two-level Otsu algorithm not only greatly reduces the times of iterations of the EM algorithm, but also effectively enhances the running speed of the algorithm in combination with downsampling in preprocessing. In addition, the energy function of the graph cut model enables accurate lake boundaries to be obtained without requiring any post-processing.

Multi-satellite cooperative observation method based on area target gridding

ZHENG Qicun, YUE Haixia, LIU Dacheng, LI Hua, REN Mingshan, JIA Xiaoxue

2024, 41 (6): 803-809. DOI: 10.7523/j.ucas.2023.019

Abstract ( 230 ) PDF (0KB) ( 0 )

By analyzing the constraints of using multiple SAR satellites to observe a specific large area target, the constraint satisfaction model is established to maximize the observation profit within a given mission time horizon. To improve the global search capabilities of the traditional tabu search algorithm, an improved tabu search algorithm with the variable neighbourhood is proposed. In implementing the algorithm with variable neighbourhood, the area target is gridded to dynamically generate observation patterns, and the observation rates are calculated for each pattern. Compared to the traditional tabu search algorithm, the variable neighbourhood tabu search algorithm proposed in this paper increases the observation profit by more than 8% while maintaining the same computational burden.

Field dynamic small object detection network based on double frame fusion

ZHAO Xiaohan, ZHANG Zebin, LI Baoqing

2024, 41 (6): 810-820. DOI: 10.7523/j.ucas.2023.008

Abstract ( 256 ) PDF (0KB) ( 0 )

Detecting dynamic small objects in complex environments in the field remains a challenging problem for defense and military applications due to factors such as more background interference in the field surveillance sensing systems, fewer pixels of small targets, and the lack of relevant open datasets. In order to solve this problem, a YOLOv5-based object detection network with double frame feature fusion (YOLO-DFNet) is proposed. Firstly, a double frame feature fusion module(D-F fusion) is introduced to process the adjacent frame features from the backbone network, calculating attention in channel, time, and space dimensions successively, to extract motion features. Secondly, a temporal trapezoidal fusion network based on an attention mechanism(TTFN_AM) is designed between the neck network and the detection head to focus on dynamic objects within receptive fields of different sizes, thereby improving the detection effect of small objects with large displacement. The experimental results on field motion small object dataset (FMSOD) show that the mean average precision (mAP) on different IoUs of the proposed YOLO-DFNet is 3.9 percentage points higher than that of YOLOv5, and also outperforms other object detection models such as Tph-YOLOv5 and YOLOv7.

Detection method and characterization of ramp events of wind speed and wind power based on swinging door algorithm

LIANG Zhi, ZHANG Zhe, SHI Yu, LIU Lei

2024, 41 (6): 821-829. DOI: 10.7523/j.ucas.2023.014

Abstract ( 315 ) PDF (0KB) ( 10 )

The ramp event of wind speed is a large increase or decrease in wind speed within a short period, causing a significant change in wind farm power, affecting the safe operation of the grid and even triggering accidents such as frequency reduction and voltage collapse. This paper selects the simultaneous data of wind turbines and meteorological towers in wind farms, identifies the ramp events by the swinging door algorithm (SDA), analyzes the duration, magnitude and change rate of the ramp events, and discusses the influence of mountainous terrain on them. In this paper, the recognition algorithm of the ramp event of wind speed and power is designed based on the SDA, and the algorithm parameters are set as follows: the time threshold 4 h, wind speed threshold 6 m·s^-1, and power threshold 1 000 kW. For the recognition of ramp events in other wind turbines, this paper suggests using 2/3 value of the difference between rated wind speed and cut-in wind speed as the wind speed threshold parameter, and 2/3 value of rated power as the power threshold parameter. The terrain influence on the ramp event is significant, and the ramp event is more related to the altitude and average wind speed at the turbine, and the time proportion of the ramp event under different terrain ranges from 6.5% to 9.8%, with the average value of 7.8%.

A framework of graph classification with self-supervised heterogeneous graph neural network

YUAN Ming, ZHAO Tong

2024, 41 (6): 830-841. DOI: 10.7523/j.ucas.2023.048

Abstract ( 253 ) PDF (0KB) ( 0 )

Graph data exists widely in various forms, and the tasks of graph classification have great significance for many problems. However, the tasks of graph classification still face many challenges, including how to make full use of the semantic information contained in the graph structures, and how to further reduce the computational complexity and the cost of obtaining labels. In this paper, a construction method of a hyper-node heterogeneous network is proposed for the first time, along with a new framework, GChgnn, which can be applied to graph classification. The GChgnn framework achieves the following goals through the introduction of a double-view graph representation mechanism and self-supervised contrastive learning: 1)measuring the similarity between the objectives of the large-scale graph classification tasks; 2)inspired by the graph matching methods, improving the accuracy of similarity measurement by the cross-graph idea, and making up for the lack of the explicit expression of graph embedding; 3)avoid the need to design complicated convolution and pooling operators in the network. After testing on some public datasets, the framework outperforms other existing methods.

Vulnerability exploitability assessment method based on network environment

ZHENG Jinghua, KAI Shaofeng, SHI Fan

2024, 41 (6): 842-852. DOI: 10.7523/j.ucas.2023.037

Abstract ( 145 ) PDF (0KB) ( 0 )

The common vulnerability scoring system is the most widely used vulnerability evaluation method, but its evaluation results tend to be the harmfulness of the vulnerability itself, ignoring the network environment factors. In view of the above problems, we propose a network environment-oriented vulnerability exploitability assessment method. Based on the experience of group experts, using statistical methods to select vulnerability attributes, the vulnerability exploitability assessment metric system is constructed. And combined with the target environment attributes, this method can evaluate the vulnerability exploitability based on the K-nearest neighbor (KNN) algorithm. This method performs accurate and intelligent assessment of known and unknown vulnerabilities, integrating the impact of the target environment and reducing the reliance on expert experience. At last, we validate the method through experiments. Our method provides a scientific decision-making basis for network security protection measures.

Current Issue