摘要:In the current air defense and anti-missile scenario, the target penetration means are complex and diverse. In the intensive multi-target tracking, the radars are susceptible to electromagnetic interference, group target occlusion, terrain occlusion, etc., and the association of multiple radars with the same target track is prone to the problem of mixed batch and batch change, which leads to the decline of the consistency of target association and the reduction of the quality of multi-radar fusion intelligence. To solve the above problems, this paper makes use of the characteristics that the range accuracy of multi-radar is higher than the angle measurement accuracy, and proposes a target track association method based on multi-radar measurement matching. First, the least square is adopted to achieve multi-radar distance cross positioning. Then, the orientation and pitch angle of the target relative to the radar are calculated from the positioning solution, and the error distribution between the measurement and the positioning solution is calculated by statistics. It helps to judge the correlation tracking of multiple radars to the same target and improve the consistency of track association. Finally, the simulation results verify the effectiveness of the method.
摘要:3D animation can vividly and intuitively display the movement process of the target, and virtual reality simulation is the highest level of simulation. 3D animation made with virtual reality simulation technology is even more intuitive, realistic and immersive. VRML is a virtual reality modeling language for network applications, which has powerful 3D modeling and rendering functions. MATLAB function can directly drive the target generation animation of VRML file. This paper mainly studies the method of 3D animation simulation combining MATLAB and VRML. We deeply analyze the theory of VRML animation by combining the animation mechanism and realization of VRML, give the interface function and operation process based on MATLAB and VRML and apply this method to the simulation of mounting process. The generated animation is intuitive, vivid and controllable, providing experimental basis for the research of virtual manufacturing. The simulation results verify the effectiveness of the method.
摘要:The domain name system (DNS), as a critical global internet infrastructure, plays a vital role in the operation of economies and societies, as well as the development of the digital economy. Its security, availability, and autonomy remain a key concern for nations worldwide. Amid escalating geopolitical conflicts, the DNS has become the primary battleground for state-level cyber warfare, as evidenced by incidents during the U.S.-Iran, Russia-Ukraine, and Israel-Palestine conflicts, highlighting the severe politicization of the DNS. This paper provides an in-depth analysis of global DNS security risks and offers strategic recommendations for strengthening China’s DNS security framework, aimed at enhancing the resilience of the nation’s cybersecurity infrastructure.
关键词:domain name system;geopolitical conflict;recursive resolution service;cyber defense system;cybersecurity
摘要:The early diagnosis of chronic obstructive pulmonary disease is of great significance to improve the prognosis of patients. In this study, a COPD respiratory sound classification model based on fast Fourier transform and Constrained Boltzmann machine feature extraction combined with particle swarm optimization algorithm optimized random forest classifier was proposed. The experimental results show that the model performs well on several evaluation indexes such as accuracy rate, accuracy rate, recall rate and F1 score, indicating its potential application value in the early diagnosis of COPD.
摘要:The analysis of pore characteristics in rock cast thin sections has been widely applied in petroleum geology for evaluating reservoir quality. Current traditional methods, such as color threshold segmentation,exhibit limitations including low accuracy and significant manual intervention requirements when segmenting pores in cast thin section images. Meanwhile, existing deep learning neural networks for pore extraction demonstrate suboptimal performance when processing low-porosity cast images with blurred pore boundaries. To address these challenges, this study proposes an enhanced network architecture based on Segnet, incorporating an optimized attention mechanism integrated with the convolutional block attention module (CBAM).The proposed method effectively preserves spatial details in feature maps while establishing long-range dependencies across spatial dimensions.Comparative experiments are conducted using diverse cast rock thin sections commonly employed in industrial practice. Evaluation metrics including pixel accuracy (PA), mean intersection over union (mIoU), and processing time demonstrate the superiority of the proposed method over conventional approaches, confirming its enhanced precision in pore extraction tasks.
摘要:Through literature reviews and surveys, this research systematically summarizes the main policies of the country on the green development of data centers. Analyze the development achievements of the data centers by examining the power usage effectiveness, water usage effectiveness, utilization of renewable energy and energy storage, and support for the national "East Data, West Computing" strategy in 246 national green data centers. The study also discusses existing issues in the green development of data centers, such as the imperfect evaluation index system and insufficient effective supply of energy-saving and carbon-reducing technologies, these findings provide a basis for the construction and promotion of national green data centers during the "15th Five-Year Plan".
关键词:data center;green development;achievement;evaluation index system
摘要:In order to address technical pain points such as the inability to transmit real-time video surveillance at port terminals and the need for manual copying when reviewing surveillance videos, as well as the serious issues of screen distortion and lag in existing 5G video transmission, technical research has been conducted on the network architecture of video transmission. This study proposes a new network architecture that uses 5G cameras (including 5G modules) to replace 5G CPE-NVR-switches and customers' existing cameras. The video stream is directly transmitted to the base station through the 5G camera and then transmitted to the video cloud platform through UPF. The experiment shows that it can effectively reduce the packet loss rate and latency of video feedback, and ensure the playback quality of port terminal video surveillance.
关键词:video feedback;network architecture;5G camera;packet loss rate;time delay