The long-term safe operation of high-speed maglev lines depends on continuous monitoring and evaluation of track structures and surrounding environmental conditions. However, traditional manual inspection suffers from low efficiency and limited coverage, making it difficult to meet the requirements of high-frequency and high-precision detection. To address these challenges, this paper investigates the operation and maintenance needs of high-speed maglev lines and designs UAV-based inspection schemes for fine-grained track detection and wide-area environmental monitoring and further proposes a functional architecture for an intelligent edge-cloud collaborative inspection system. In addition, this study systematically analyzes the key technical challenges in implementing the proposed system, including small-object perception, motion-blur suppression, out-of-distribution target identification, and robustness under adverse environmental conditions. Future work will focus on constructing a dedicated dataset and developing key detection algorithms to support engineering implementation and real-world deployment of the system. This study provides feasible insights and technical references for advancing intelligent inspection systems for high-speed maglev lines.

The design of multistage induction coilguns is complicated by strong parameter coupling and high computational costs. Furthermore, conventional optimization methods are often inefficient and prone to converging at local optima. To efficiently explore the resulting high-dimensional design space with a multistage induction coilgun, this article proposes a design framework combining surrogate modeling with stage-by-stage Bayesian optimization (BO) to maximize peak velocity under strict sample-size and computational constraints. Specifically, the current filament model (CFM) is employed to rapidly generate sample data. A Gaussian process (GP) surrogate model is then constructed to approximate the opaque system, with parameters optimized via the lower confidence bound (LCB) acquisition function. Case studies on single- and three-stage coilguns demonstrate that applying a stage-by-stage BO strategy increases the peak velocity of the three-stage system from 148.759 to 181.99 m/s. Furthermore, the proposed framework significantly outperforms both the genetic algorithm (GA) and particle swarm optimization (PSO) in computational efficiency and global search capability. Ultimately, this work provides a robust solution for optimizing multistage induction coilguns under small-sample conditions.

To address the challenges of compromised detection accuracy caused by near-shore clutter in synthetic aperture radar (SAR) ship detection and the limited deployability of complex algorithms on embedded systems, this article proposes Lightweight-YOLOX (L-YOLOX), a lightweight SAR target detection algorithm optimized for terminal devices. First, we devise a new feature extraction module based on the MobileNetV3 block to reduce the parameters of traditional YOLOX while strengthening feature representation. Additionally, we incorporate a cross-channel local connection structure to construct an efficient lightweight feature extraction backbone, which is beneficial to improving the network's ability to fuse SAR ship target information. Next, we develop a multiscale detection block by using a feature pyramid architecture and dilated convolution to improve the network's multiscale detection performance. Finally, we integrate a lightweight convolutional attention mechanism into YOLOX's Neck structure to enhance the expression of important target detail information and propose the Alpha-AIoU loss function to optimize the gradient propagation process and the network's weight update. Ablation experimental results on the SAR Ship Detection Dataset (SSDD) dataset show that our network achieves an average precision (AP) of 90.8%, outperforming Baseline YOLOX with a 70.1% reduction in parameters and a 46.9% decrease in computational cost. Our network also demonstrates a marked enhancement in robustness, validating the effectiveness of our innovations. Some comparative experiments with other state-of-the-art algorithms on SSDD and High High Resolution SAR Images Dataset (HRSID) further confirm the advantages of our network in terms of SAR image lightweight detection performance and generalization capacity.

To address the problem of continuous launch in the case of the separation between the armature and the payload, an armature recovery type continuous launch method is proposed. This method utilizes the reverse electromagnetic force generated by the final coil to decelerate the armature, ultimately achieving automatic reset to the initial position for the next launch turn. The electromagnetic force characteristics were theoretically analyzed, and the corresponding dynamic and kinematic models were derived. Experimental results show that by adjusting the trigger timing and capacitor voltage, the payload launching and armature recovery tasks can be successfully completed at different speed levels, verifying the effectiveness of this method.

The solar sail spacecraft can reflect sunlight to obtain light pressure thrust, and can also adjust the reflectivity of the sail surface to obtain attitude adjustment torque to realize the light pressure attitude control function. Firstly, a Reflectivity Control Device (RCD) is developed for a hexagon solar sail spacecraft equipped with Liquid Crystal Membrane (LCM) array, and the differential drive signal with half-cycle pulse width modulation is used to ensure the zero-crossing alternating and isochronous drive characteristics required for liquid crystal membrane control, and effectuate the multi-state and multi-channel control of the reflectivity parameters of the sail surface. Secondly, two ground testing systems are developed in proportion, and the test results indicate that the RCD can consummate the multi-state regulation of the reflectivity parameters of membrane array in multi-channel way, demonstrating its potential as an effective solution for advanced solar sail applications.

This study focuses on the audience's emotional perception of AIGC (Artificial Intelligence Generated Content) films. The dual attributes of film as an art and a medium play an important role in human perception and cognition of reality, and generative AI is gradually penetrating into all aspects of film production, bringing about new breakthroughs in image production, special effects generation, and so on. This study selects representative AIGC short films as the research object, aiming to explore the audience's understanding of the AI elements in AIGC films, the degree of awareness and acceptance of AIGC films by audiences from different backgrounds, as well as the factors affecting the differences in audience perceptions by means of a quantitative research method. This study is of great significance in understanding how viewers perceive AIGC films, especially their reaction to the AI elements in the films, and helps to promote the development of the film and television industry in the new era of digital civilization.

Due to traditional recovery systems lacking visual perception, it is difficult to monitor UAVs’ real-time status in communication-constrained or GPS-denied environments. This leads to insufficient ability in decision-making and parameter adjustment and increase uncertainty and risk of recovery. Visual inspection technology can make up for the limitations of GPS and communication and improve the autonomy and adaptability of the system. However, the existing RT-DETR algorithm is limited by single-path feature extraction, a simplified fusion mechanism, and high-frequency information loss, which makes it difficult to balance detection accuracy and computational efficiency. Therefore, this paper proposes a lightweight visual detection model based on transformer architecture to further optimize computational efficiency. Firstly, aiming at the performance bottleneck of existing models, the Parallel Backbone is proposed, which captures local features and global semantic information by sharing the initial feature extraction module and the double-branch structure, respectively, and uses the progressive fusion mechanism to realize the adaptive integration of multiscale features so as to balance the accuracy and lightness of target detection. Secondly, an adaptive multiscale feature pyramid network (AMFPN) is designed, which effectively integrates different scales of information through multi-level feature fusion and information transmission mechanism, alleviates the problem of information loss in small-target detection, and improves the detection accuracy in complex backgrounds. Finally, a wavelet frequency–domain-optimized reverse feature fusion mechanism (WT-FORM) is proposed. By using the wavelet transform to decompose the shallow features into multi-frequency bands and combining the weighted calculation and feature compensation strategy, the computational complexity is reduced, and the representation ability of the global context is further enhanced. The experimental results show that the improved model reduces the parameter size and computational load by 43.2% and 58% while maintaining detection accuracy comparable to the original RT-DETR in three datasets. Even in complex environments with low light, occlusion, or small targets, it can provide more accurate detection results.

Short distance recovery of Unmanned Aerial Vehicle (UAV) based on eddy current braking is a newly developed technology. The UAV is arrested by an arresting cable with eddy current brake, and its kinetic energy is converted into heat through eddy current effect. However, due to the fact that eddy current brake for aircraft is mostly large ground-fixed facility, there is currently no suitable analytical model for optimizing the design of eddy current brake deployed on mobile platform. On the basis of the model of arranging magnetic poles along the disk surface, the analytical models of eddy current brake with magnetic poles arranged along the circumference and their combination are provided based on the equivalent magnetic circuit method, the skin effect and eddy current demagnetization effect in this paper. The results indicate that those models have good accuracy. Due to the fact that analytical models consume less computational time than three-dimensional finite element methods, they can serve as effective tools for optimizing design.

Carriage crowd density monitoring is a key component in developing intelligent transportation systems, such as maglev transportation system. Surveillance images captured by sensors, such as carriage monitoring cameras, serve as a new solution for estimating crowd density inside the carriage due to their wide coverage and real-time updates. In this study, a passenger head detection dataset (PHD) is developed using 3717 images acquired from carriage surveillance. Based on these images, over 67,215 head instances are precisely annotated manually. To address the issue of insufficient feature fusion in existing detection algorithms, an efficient cross-scale feature enhancement (CFE) module is proposed and introduced into the advanced YoloX model. The PHD dataset is, to the best of our knowledge, the first public dataset of surveillance images for carriage crowd density estimation. To prove the usability of the PHD dataset and the validity of the proposed method, 12 different versions of detectors are applied and compared. The results demonstrate the performance of these algorithms in the detection of passenger heads. Our research offers a new approach for carriage crowd density estimation. The dataset is publicly available at: https://github.com/Xujiajing111/PHD.

Solar sails spacecraft can achieve long-duration, propellant-less flight by utilizing solar radiation pressure thrust. A reflective control device (RCD) based on liquid crystal membrane can alter the distribution of solar radiation pressure on the sail surface, enabling adjustable torque that can be utilized for reaction wheel (RW) unloading in deep-space environments with weak magnetic fields. Current research on RW unloading control for solar sails predominantly employs pure numerical simulations based on ON/OFF reflectivity control, which struggles to incorporate the nonlinear characteristics of physical RCD devices and RWs. Firstly, a novel continuously adjustable reflectivity control device is introduced to conduct modeling and numerical simulations of a RW unloading control system based on continuously adjustable light pressure torque. Secondly, a Hardware-in-the-Loop (HIL) simulation system is constructed using self-developed RCD device and off-the-shelf RW, verifying the effectiveness of the novel continuously adjustable RCD in achieving momentum unloading and revealing the nonlinear characteristics of the RW unloading speed curves.

Quasi-2D microsatellites possess the characteristics of high launch efficiency, low development cost, and low-drag flight capability, and are suitable for future low earth orbit (LEO) and very low earth orbit (VLEO) space missions. In order to adapt to the stacked launch mode, the quasi-2D microsatellite adopts a new high-functional density design, which leads to a new mechanical and thermal environment problem. Taking the specific platelike configuration as an example, the structure and thermal control design of the quasi-2D microsatellite are firstly carried out, and the mechanical/thermal modeling and simulation of the satellite are carried out, and the fundamental frequency and temperature distribution characteristics of the structure are obtained, which provides a data reference for the subsequent design. Secondly, the satellite structure was rapidly developed through aerospace additive manufacturing, and the preliminary thermal design scheme of passive heat dissipation of high thermal conductivity silicone rubber sheet was adopted. Through thermal simulation modeling and physical testing, it has been proven that the new quasi-2D satellite has uniform transient heat distribution and controllable temperature during continuous operation. The new method meets the requirements of fast response and low cost of low-orbit missions, and provides a new solution for the efficient completion of future LEO and VLEO space missions.

Aiming at the wearing adaptability and comfort of energy harvesters, an energy harvesting device based on magnetic springs and friction power generation is designed to collect human movement energy. The effects of magnetic energy product, magnetization direction and winding height of permanent magnets on the performance of magnetic spring power generation were studied, the influence of friction position on friction power generation was studied, an energy management circuit was designed, and a wearable composite power supply system was made, which improved the adaptability of magnetic springs and friction power generation in the wearable field, and provided a reference for the application of energy harvesting technology in the wearable field.

高速磁浮列车是利用电磁力实现车辆与轨道无接触高速运行的一种新型交通工具,车辆的导向和制动性能受到轨道导向不平顺的影响.为了保证高速磁浮车辆运行的安全性、稳定性和舒适性,设计一种结构简化、低成本和搭载式的磁浮轨道导向不平顺检测系统.该系统基于惯性基准法原理实现检测,由加速度计、测距传感器、数据记录仪和里程检测模块组成,并未使用陀螺仪和倾角仪测量载体的姿态角变化.分析了车辆姿态变化对导向不平顺检测误差的影响,因未修正姿态导致的检测误差绝对值在直线段轨道达到0.4 mm,而在曲线段轨道超过了3 mm.为了降低缺乏姿态观测所致误差,提出一种设计线型辅助的策略用以部分替代倾角仪功能,即以列车所在位置轨道的横坡角和纵坡角分别近似替代载体的侧滚角和俯仰角低频分量,并用于补偿加速度积分中的重力和离心力分量,仿真表明该方法可将曲线段轨道的检测误差降低至0.6 mm.此外,结合磁浮轨道刚度大、变形小以及分段铺设的特点,利用分段直线拟合方法对不平顺检测结果进行平滑处理,从而进一步降低缺乏姿态观测的影响,保证系统具有足够的检测精度.通过小车检测试验,结果表明所设计系统及数据处理方法可实现±0.5 mm之内的检测误差.

Currently, the target tracking algorithm based on discriminative correlation filter (DCF) and deep learning (DL) plays an increasingly important role in UAV (unmanned aerial vehicle). However, existing algorithms have limitations such as limited search region, difficulty in re-capturing targets when tracking is lost, and high computational complexity, which makes them difficult to apply to the UAV platform. In this paper, a target re-detection tracking algorithm (RDT) after tracking loss is designed for mobile platforms. RDT uses DCF in simple scenes to track targets and establish a target model between different frames. To ensure the algorithm when tracking is lost due to occlusion or moving out of view, an efficient switching criterion is designed to indirectly invoke the detection algorithm to re-capture the target and guide online learning of DCF. And the experiment results on the OTB benchmark show that RDT can re-capture the target after tracking is lost, and the speed of operation on the CPU is 85fps.

Aim to improve the power density of the electromagnetic ejection system of UAV, the finite control set model prediction is adopted as the control strategy from the perspective of improving the efficiency. The semi-active control of hybrid energy storage system and the drive control of ejection motor are considered together. According to the different requirements of commutation and non-commutation, the finite control set of the electromagnetic ejection system is designed, control optimization is carried out from the system. Simulation results show that the proposed control strategy can effectively reduce thrust fluctuation, stabilize bus voltage, reduce switching loss and improve the efficiency of electromagnetic ejection system.

The high-speed maglev train is a new type of transportation. The long stator track plays a critical role in the levitation guidance and traction system. Therefore, its condition directly affects the operation of maglev trains. It is extremely important to detect the abnormal condition of high-speed maglev tracks to ensure the stable, safe, and reliable operation of the train. In this article, an onboard image detection system is designed for high-speed maglev tracks, which can accurately obtain the image of long stator tracks under the harsh conditions of limited installation space, insufficient illumination, and rapid operation of vehicles. High-speed maglev trains are not yet in widespread use. In China, there is currently only one demonstration operating line located in Shanghai, and the length of the track test line is limited. Therefore, the number of track images that can be acquired is extremely limited. In view of the lack of defective samples of high-speed maglev tracks, this article proposes a data enhancement method based on sample generation and image fusion to augment the dataset of defective samples. To improve the quality of generated high-speed maglev track defect images, a joint attention layer (JEA) combining squeeze-and-exception (SE) block and spatial attention module (SAM) is designed and introduced into the generative adversarial network (GAN). This work provides a data basis for the study of track defect detection of high-speed maglev trains. In addition, this article detects the defects of high-speed maglev tracks via deep learning-based target detection algorithms, which can automatically detect, accurately classify and locate the defects of stator surface and cables, filling the gap in the field of high-speed maglev track defect detection.

A scheme of detecting magnetic field anomaly is proposed to find local short circuit faults in long stator cores of high speed maglev transit. And a composite magnetic anomaly signal processing method is designed based on wavelet with fractal, screening out a little anomalous data from massive data. First, suspicious data are quickly recognized by applying multiple fractal spectra, then being confirmed by wavelet modulus maxima calculation. The experiment results show that the method is immune to the fluctuation of 8–12 mm suspension gap and its hit rate can be improved by more than 66%.

常导磁浮列车采用无接触的悬浮和直线牵引原理,具有无磨损、低噪音、高速、安全、平稳、舒适等优点,受到了广泛关注.按照速度,常导磁浮列车可分为高速磁浮列车、中速磁浮列车和中低速磁浮列车等,本文主要针对高速磁浮和中低速磁浮两种形式的车辆传感器展开研究讨论.

为了实现陆基无人机电磁弹射器高机动性及其直线弹射电机的高功率密度,针对动圈式永磁直线直流电机,提出精英保留的多种群遗传算法(Multi-Population Genetic Algorithm with Elite Retention,MPGAER)的电机最大功率密度优化方法.以磁通密度和电流密度为约束条件,利用其搜索能力强、收敛速度快的特点优化电机的结构参数,并与磁路法初始设计结果和传统遗传算法优化结果进行比较.结果表明:与磁路法初始设计相比,MPGAER能使电机质量减少6.25％,功率密度提高10％,电机动态性能得到提高;MPGAER优化设计的电机功率密度高于遗传算法设计结果,所提方法有效地解决了在优化过程中出现易收敛于局部最优点和寻优效果差的问题.

A sensorless commutation control method based on the zero crossing principle of line back EMF is proposed for permanent magnet DC linear motor with distortion of opposite EMF. Through qualitative analysis of waveform and quantitative derivation of series theory formula, it is proved that even if the opposite potential is distorted, the zero point of line back potential is still consistent with the actual electronic commutation point of the system. At the same time, in order to avoid missing the zero point of line back EMF, a commutation control method based on s function is constructed. The simulation results show that the method meets the system requirements.