Nitrogen dioxide (NO2) in ship exhaust is one of the major atmospheric pollutants, and its efficient monitoring is critical for both maritime emission control and atmospheric environmental management. Current monitoring methodologies exhibit deficiencies in spatiotemporal resolution and operational efficiency, failing to meet accurate regulatory requirements. This study proposes a UV remote sensing imaging system with high spatiotemporal resolution for ship exhaust NO2 monitoring. By analyzing the ultraviolet absorption characteristics of NO2 in the 440 nm band, a dual-channel architecture is designed to separate background interference and extract pure NO2 absorption signals. Integrated with a light dilution correction model, the system achieves high-precision inversion of NO2 column concentrations. Meanwhile, a machine vision optical flow algorithm is introduced, and a real-time estimation method for NO2 emission rate is constructed by quantifying the correlation between the dynamic diffusion process of the plume and the spatiotemporal concentration gradient. The experimental results show that the NO2 emission rates of the two ro-ro passenger ships measured by the system have a significant linear relationship with the ship speed, with goodness-of-fit values of 0.922 and 0.932 respectively, which verifies the effectiveness and feasibility of the system in real-time monitoring of ship exhaust emissions. Error analysis reveals that light dilution correction reduces the relative errors of emission rates by approximately 6 % and 1 % respectively, enhancing robustness in long-range monitoring. This research establishes innovative technological solutions for precision NO2 monitoring in ship emissions, advancing the deep integration of optical remote sensing and machine vision technologies in environmental monitoring. It provides crucial scientific foundations for evaluating ship emission reduction effectiveness and formulating atmospheric pollution control policies.
To address the limitations of existing algorithms in segmenting and detecting occluded objects, which often suffer from misdetection and omission errors, this paper proposes a novel occluded object segmentation model based on a Duplex Residual Network (DRNet) combined with Efficient Intersection over Union (EIoU). First, DRNet replaces the traditional residual neural network by employing fewer Batch Normalization and Rectified Linear Unit layers, while enhancing the extraction of image receptive field features through strategically designed, larger-sized depthwise separable convolutions. Second, the Clus-ter EIoU Non-Maximum Suppression algorithm is introduced to replace the conventional Non-Maximum Suppression (NMS) algorithm. This modification effectively incorporates geometric factors into the suppression process, reducing excessive sup-pression of bounding boxes for occluded objects. Third, the EIoU loss function is adopted to replace the original Smooth L1 loss, accelerating network convergence and improving bounding box regression accuracy. The proposed model is pre-trained on the publicly available MS COCO datasets and evaluated on multiple occlusion datasets with varying degrees of object over-lap. Experimental results demonstrate significant improvements over the baseline Mask R-CNN network, with the proposed algorithm achieving gains of 1.7% in Box AP and 1.3 % in Mask AP on the COCO validation set. Furthermore, the model substantially improves bounding box detection accuracy and mask segmentation accuracy for heavily occluded objects. These results confirm the effectiveness of the proposed method for occluded object segmentation tasks.
Ultrawideband (UWB) technology has garnered considerable research interest for indoor positioning applications owing to its temporal resolution and strong signal penetration capabilities. However, the traditional classification of line-of-sight (LoS) or non-line-of-sight (NLoS) channel condition fails to characterize the complex indoor environments, and the existing models are trained on the entire dataset without considering the heterogeneity of the data, resulting in suboptimal positioning accuracy and reliability. To overcome these limitations, we propose an innovative autonomous classification method called SimCLR-CIR-SC that combines contrastive learning principles from the SimCLR architecture with spectral clustering (SC) techniques for enhanced feature extraction from channel impulse response (CIR) measurements. Building on the results of autonomous classification, we develop a temporal convolutional network with attention (TCN-A) architecture to discriminate between diverse channel state categories. For each identified channel condition, a dedicated TCN-A model is deployed to predict ranging errors, which subsequently perform distance calibration through error compensation and adaptive weight assignment in the weighted least-square (WLS) positioning algorithm. Experimental evaluations reveal that the proposed SimCLR-CIR-SC method achieves superior autonomous channel state classification and labeling performance compared to three clustering algorithms. Notably, the TCN-A classification model attains an accuracy of 98.16%, surpassing five existing classification models. Furthermore, the proposed positioning method achieves an average error of 0.57 m with three anchors, enhancing the positioning accuracy by at least 31.3% compared to four benchmark methods. Obviously, the positioning accuracy can be further improved as the number of anchors increases, and the average error is 0.278 m with seven anchors.
A novel wavelength selection algorithm,based on wave cluster interval,for infrared spectroscopy in the detection of volatile organic gases is presented.The algorithm employs a series selection mode,utilizing characteristic wavelength point cluster classification and absorption peak interval screening.To begin with,cluster analysis is conducted to retain prominent absorption peak features while minimizing the potential for algorithmic over splitting or random uncertainty in wavelength intervals.Subsequently,an improved moving window method is devised,and a greedy algorithm is employed to re-screen wavelength points within the same cluster class.This process ensures the retention of the optimal wavelength range,crucial for representing spectral characteristics and facilitating subsequent model predictions.Experimental validation was conducted using infrared spectral data of styrene,para-xylene,and o-xylene,employing four models:partial least squares,ridge regression,support vector machine.The results demonstrate that,while maintaining model accuracy,the dataset can be reduced to 43.71％～36.35％of its original size.Additionally,utilizing a dataset comprising three gases(two concentrations each),as well as fully arranged and combined mixed gases,we conducted comparative experiments on three different CNN structures.The effectiveness of the proposed algorithm in reducing machine learning model complexity while ensuring prediction accuracy was validated through experimental comparisons before and after spectral waveform selection,with the CNN prediction models demonstrating a 90％increase in operational efficiency post-wavelength selection.
A complete communication system must encompass a full two-way communication link,which includes both uplink and downlink.The uplink has always been a challenge for two-way communication systems.In recent years,the rapid development of visible light communication technology,with its advantages of no electromagnetic radiation,large communication capacity,and environmental friendliness,can serve as a supplement to traditional uplink solutions.The article first introduces the application scenarios and system composition of visible light communication,and then provides a review of the current research status of visible light uplink at home and abroad in recent years.In addition,it presents various schemes for visible light uplink,such as visible light with radio frequency,visible light with visible light,visible light with power line carrier,and single-source reverse modulation technology.Finally,it summarizes the current issues faced by the visible light communication uplink and summarizes the advantages and disadvantages of various schemes,as well as prospects for future development trends.
Research questions:In optical wireless communication systems,atmospheric turbulence can cause the transmission beam to expand,drift and light intensity fluctuation,which will seriously reduce the signal quality of the receiving end and reduce the performance of the communication system.Therefore,the study of methods to suppress atmospheric turbulence is the key to improve the performance of optical wireless communication systems.Method and process:Large-aperture receiving technology,diversity technology,partially coherent beam technology and adaptive optics can effectively suppress the atmospheric turbulence effect,which is an important means to improve the performance of optical wireless communication systems.Detailed detail the principle of suppress atmospheric turbulence and its means.These key technologies can improve the quality of the received signals and enhance the reliability of the communication system by changing the transmission or reception strategy,regulating the structure of the optical field,enlarging the receiving aperture,and compensating for wavefront distortion.Meanwhile,the effects of different parameter indicators on the system performance are also analyzed.The current status of domestic and international research on the relevant suppression techniques is discussed,and the improvement of different performance indexes of the system under the influence of atmospheric turbulence by the relevant techniques is showed.Conclusions:Finally,the challenges and problems in atmospheric turbulence suppression in the field of optical wireless communication are summarized,and the future development trend of the technology is outlooked,which can provide a reference for the future development in this field.
In order to obtain a portable and simple vibration energy harvester and overcome the problem of low output of traditional energy harvesters,a double-L bracket kickback piezoelectric energy harvester was proposed,which combined the nonlinear characteristics of the tensile structure with piezoelectric technology,which could effectively improve the dynamic response and output performance of the energy harvester.The effects of external resistance,magnetic distance,excitation acceleration and angle on the output performance of the harvester were analyzed.The experimental results show that the output power reaches the peak value of the harvester at the optimal external resistance value,and the optimal external resistance value is 200 kΩ.The introduction of magnetic force can significantly improve the output performance of the energy harvester.At a magnetic distance of 18 mm,the harvester captures the most energy,and its optimal output performance reaches 1.12 mW at 12.1 Hz.In addition,the excitation acceleration has an obvious impact on the output characteristics of the energy harvesting system,and the output voltage and output power of the energy harvester also increase with the larger the excitation acceleration,and the maximum output power of the energy harvesting system reaches 1.39 mW under the condition of the excited acceleration of 0.4 g and the frequency of 12 Hz.The harvester has good output voltage and output power in the angle range of 0°～45°,and has the advantage of working under the condition of uncertain excitation direction.Practical application experiments further prove that the energy harvester can continuously output a large and stable voltage,which provides an effective solution to solve the problem of low output of traditional energy harvester.
Fiber optic interferometer has the advantages of small size, light weight, anti-corrosion, anti-electromagnetic interference, high sensitivity, etc., and is widely used in the measurement of temperature, humidity, magnetic field and other parameters. In recent years, researchers have dramatically improved the measurement sensitivity of interferometric fiber-optic sensors by cascading or paralleling two fiber-optic interferometers to produce an optical Vernier effect. When the free spectral ranges of the two fiber optic interferometers are close but not equal, the resulting Vernier effect is called the normal Vernier effect, when the free spectral range of one fiber optic interferometer is about an integer multiple of the other fiber optic interferometer, the resulting Vernier effect is called the harmonic Vernier effect. In this paper, a parallel optical fiber temperature sensor based on two Sagnac interferometers is proposed, where the interferometers SI1 and SI2 are connected to the two outputs of the fiber-optic coupler C3, where SI1 is a reference interferometer and SI2 is a sensing interferometer. When the length of the Panda fiber in the SI2 interferometer is approximately i+1 of the length of the Panda fiber in the SI1 interferometer (i is 1, 2, 3 center dot center dot center dot) times, the two interferometers will produce an i-order harmonic Vernier effect. When i is 0, it produces an normal Vernier effect, at which time there will be a single envelope in the interference spectrum. When i is 1, it produces a first-order harmonic Vernier effect, at which time there will be a double envelope in the interference spectrum. When i is 2, it produces a second-order harmonic Vernier effect, at which time there will be a triple envelope in the interference spectrum. In other cases, the order is analogous. We have numerically simulated the theoretical analysis and the free spectral range of the interferometric spectrum of the length interferometer SI1 with a Panda fiber of 520 mm at constant temperature is 9.13 nm. The SI2 interferometers with lengths of 572 mm, 953 mm, and 1 430 mm of the Panda fiber have free spectral ranges of 8.30 nm, 4.98 nm, and 3.32 nm, respectively. When the temperature is increased from T-0 degrees C to T-0+1 degrees C, the interference spectra of SI2 interferometers with different Panda fiber lengths are all shifted to the short-wave direction, and the shifts are all about 1.89 nm, which is consistent with the theoretical analysis. The parallel interference spectra of interferometer SI1 and SI2 with Panda fiber lengths of 572 mm, 953 mm and 1 430 mm show single, double and triple envelopes respectively, indicating that the two interferometers produce the normal Vernier effect, first-order and second-order harmonic Vernier effects, respectively, and from the theoretical calculations. It can be seen that the amplification of the normal Vernier effect, first-order and second-order harmonic Vernier effects are all 11 times. When the temperature increases from T-0 degrees C to T-0+1 degrees C, the single envelope moves in the short-wave direction, while the double and triple envelopes both move in the long-wave direction, which is opposite to that of the single SI2. In addition, the shifts of the single, double and triple envelopes are all about 20.7 nm due to the fact that the Vernier magnification is the same for the normal Vernier effect, first-order and second-order harmonic Vernier effects. It is experimentally concluded that the interference spectra of SI2 are blueshifted in the temperature range from 40 degrees C to 50 degrees C, and the shifts are all about 1.89 nm, which is consistent with the theoretical analysis and simulation results. The temperature sensitivity of the sensor corresponding to the normal Vernier effect is -20.67 nm/degrees C, the temperature sensitivity of the sensor corresponding to the first-order harmonic Vernier effect is 21.34 nm/degrees C, the temperature sensitivity of the sensor corresponding to the second-order harmonic Vernier effect is 21.18 nm/degrees C, and the fiber optic sensors corresponding to the harmonic Vernier effect and the normal Vernier effect have almost the same temperature sensitivities, which are both about 21 nm/degrees C. This results are consistent with the theoretical analysis and simulation results. The above experimental results show that the temperature sensitivity of the SI2 interferometer is independent of the length of the Panda fiber, although the magnification is the same, the harmonic Vernier effect and the normal Vernier effect correspond to the detuning of the length of the Panda fiber are obviously different, the detuning corresponding to the normal Vernier is 52 mm, and the detuning corresponding to the first-order and second-order harmonics is -87 mm and -130 mm, respectively. This shows that the higher the order, the larger the detuning amount, which is approximately a multiple increase. The above experimental results are consistent with the theoretical analysis. Since the larger the detuning amount, the easier the Vernier magnification can be controlled and realised, the harmonic Vernier effect is obviously superior to the normal Vernier effect from the preparation point of view. This study can provide an important reference for the subsequent study of optical Vernier effect.
飞行器中继可以快速构建端到端的通信网络，增强通信系统的覆盖范围和传输能力，强化信号强度和提高数据传输速度，提高通信系统的可靠性。总结了无人机中继技术以及无人机中继在无线光通信中的国内外发展现状，并对无人机中继在无线光通信中关键技术进行了深入研究。对信道特性及其对通信的影响和抑制技术进行了分析，比较了不同调制解调方法的优缺点和通信过程中的光束捕获，跟踪和对准的具体方法，详细介绍了光束的快速对准。最后展望了无人机中继在无线光通信中的发展方向。
Optical wireless communication is an important means of modern communication,with rich spectrum resources,anti-electromagnetic interference and other advantages,can be used as an important supplement to traditional radio frequency technology,which is expected to provide an important technical driving force for future industrial manufacturing.This paper firstly introduces the relevant applications of optical wireless communication under the background of industrial Internet,and summarizes the research status at home and abroad in recent years,summarizes the typical research progress,expounds the channel model and key technologies of optical wireless communication system,and introduces the RF/visible light heterogeneous technology used to ensure the reliable transmission of the uplink and downlink of the communication system.Finally,the paper summarizes the current problems faced by optical wireless communication in the industrial Internet,and looks forward to the future development trend,which can provide reference for the future research and development of optical wireless in this field.
棒形复合绝缘子的爬电距离是其质量管控中的重要关注点.针对现有测量方法依赖人工,效率较低且误差大,提出了一种基于机器视觉的复合棒形绝缘子爬电距离测量方法.首先,设计了棒形复合绝缘子图像采集平台,获得了其轴向连续图像.采用SIFT算法对获取到的连续图像进行拼接,通过Canny边缘检测算法构建识别模型,实现了棒形悬式复合绝缘子的爬电距离的自动测量.试验表明,图像测量方法的重复测量标准差约为传统方法的5％～12％,且其测量效率较传统方法提高6倍以上,实现了棒形复合绝缘子爬电距离准确高效的测量.
弧齿锥齿轮作为收获机主动力输出的关键零部件,其故障表现通常为激励脉冲,为实现农业收获机主传动齿轮箱故障及时有效地监测诊断,本文提出基于物理模型驱动优化的小波包分解方法(wavelet packet decomposition,WPD).针对齿轮损伤的多分量调制现象,该方法根据小波基函数特定时频窗口分析信号的特点,通过建立齿轮损伤集中参数模型,辅助筛选适应齿轮损伤特性的小波包分解系数,以此优化分解信号所选用的小波基函数,使之具有更好的提取齿轮故障特征信息的能力.通过对实验信号和藠头收获机齿轮故障信号的包络谱分析,验证了该方法能够有效地应用于收获机齿轮故障诊断.
随机配置网络(SCNs)具有通用逼近能力和快速建模特性,已成功应用于大数据分析.在SCN的基础上,块增量随机配置网络(BSC)使用块增量机制提高训练速度,但增加了模型结构的复杂程度.为了解决上述难题,提出遗忘因子随机配置网络(FSCN-Ⅰ和FSCN-Ⅱ)驱动的自适应切换学习模型(ASLM).该模型利用正态分布配置隐含层节点的输入参数.FSCN-Ⅰ通过误差值和遗忘因子调整节点块的尺寸,提高训练速度.FSCN-Ⅱ引入节点移除机制降低模型结构的复杂程度.ASLM由FSCN-Ⅰ和FSCN-Ⅱ构成,两者根据自适应变化的边界随机切换以提高模型的训练速度,并在FSCN-Ⅰ的基础上降低模型结构的复杂程度.最后,通过基础数据集和工业实例,表明该方法的有效性.
准确高效地对人体姿态进行语义描述是识别人类行为的重要部分,也是快速了解个体状态以及发生事件的关键.近年来,人体关键点检测技术获得了长足的发展,然而针对人体姿态语义描述的研究并未引起足够重视.为此,本文提出了一种基于几何统计的人体姿态语义描述方法.首先将获得的人体关键点划分为若干集合,然后提取每个关键点集合的几何分布特征用于描述人体姿态,最后采用层次策略判断人体姿态的语义.该方法采用了集合的思想来提高识别人体姿态的鲁棒性.在不同真实场景数据集上的实验结果表明,所提方法在简单和复杂单人姿态的IFD和PASCAL数据集上识别人体姿态的平均准确率分别达到了 90.8％和77.1％,对于复杂多人姿态的MPⅡ数据集准确率为77.2％,均优于对比方法,可见所提方法在关键点缺失等情况下依然能够实现较准确的人体姿态语义描述.
磁性材料是一种重要的刺激响应材料,可通过外部磁场穿透组织和器官实现无线远程控制,具有生物兼容性高、磁场控制简单和调控速度快等特点,广泛应用于医疗机器人、人造器官、生物化学合成和药物递送等生物医学领域.复杂的工作场景和多功能需求对磁性材料的精确控制提出了更高的要求,从基于磁力的简单平面驱动,到基于磁力和磁力矩的复杂空间驱动;从基于材料本身运动变形进行功能实现到作为多功能柔性电子器件载体实现更复杂的环境检测等.本文介绍了几种常用磁性材料的磁学特性、磁场控制平台和磁极编程技术,并展示了磁性液体、磁性块体和磁性薄膜3种不同形态磁性材料在生物医学领域的应用进展和面临的诸多挑战,最后对未来的发展趋势进行了展望.
非金属管道在地下管网建设中得到广泛应用.探地雷达法作为地下公共设施主要检测方法之一,常被用于地下非金属管线探测中.由于探测过程中存在非金属管道回波信号微弱、地下结构复杂、波速难以精确估算等问题,使雷达探测成像及定位成为难题.针对上述问题,研究了适用于埋地非金属管道的波速估算和回波信号处理方法,提出了空间频域插值成像和二值化定位算法;通过数值仿真验证了所述成像定位方法的可行性,并进行了 0.400、0.600和1.100 m埋深的非金属管道现场测试.实验结果表明,所提方法能够有效消除双曲线效应,探测区域的成像定位误差分别为0.006、0.006和0.012 m,满足埋地非金属管道定位需求.研究成果可为埋地非金属管道成像及定位提供技术支持,有益于提高探地雷达对地下非金属管线的探测精度.
针对传统可见光在黑暗环境中难以实现人员行为检测与身份识别的问题,本文结合红外热成像技术基于百度飞桨深度学习框架研究了一种面向黑暗环境的人员行为检测与身份识别算法。首先经过实地采集,自主构建红外热成像人员行为数据集总计10 900张9种行为类别以及双光人脸数据集总计3 000张30位人员。针对行为检测方面,基于轻量化网络PP-LCNet改进YOLOv5骨干网络进行人员行为检测,大幅度减少模型参数并提高检测精度与推理速度。针对人脸识别方面,引入Cycle GAN算法改进Insight Face实现将红外人脸转化为可见光人脸进行身份识别,提高在黑暗环境下人脸识别准确率。最后实现红外人员行为检测网络与人脸识别网络的级联工作,在黑暗环境下可以实时行为检测与身份识别,具有很好的应用效果。实验结果表明,基于PPLCNet轻量化改进的YOLOv5相对于原网络模型参数减少56.4%,平均精度m AP由89.1%提高至94.7%,推理速度由68提高至101 fps;基于Cycle GAN算法改进Insight Face相对于原网络黑暗环境下识别准确率由84%提高至99%。
针对风电齿轮箱故障预警中数据信息挖掘不充分问题,提出一种基于图注意力和时间卷积网络的风电齿轮箱故障预警方法.分别从时间与空间尺度建立各特征点的物理联系,拓宽特征维度以提升故障预警精度.图注意力网络构建不同数据测点间的空间拓扑结构,遍历每个节点的相邻节点进行加权求和达到聚合信息的目的;时间卷积网络使用特殊的因果膨胀卷积和残差网络,扩大感受野,提升时间特征捕捉能力.以华北某风电场实际数据为例进行验证,结果表明,提出方法能够在故障发生前122 h监测到风电齿轮箱的异常状态并发出预警信号;与其他方法进行对比,提出方法预警时间提前52～63 h,模型预测误差减小1.05％～3.76％;使用t-SNE和概率密度曲线提升结果可解释性.
