Combining machine learning algorithms with traditional methods for resolving the atomic-scale dynamic structure of monolayer MoS_(2) in high-resolution transmission electron microscopy

High-resolution transmission electron microscopy(HRTEM)promises rapid atomic-scale dynamic structure imaging.Yet,the precision limitations of aberration parameters and the challenge of eliminating aberrations in Cs-corrected transmission electron microscopy constrain resolution.A machine learning algorithm is developed to determine the aberration parameters with higher precision from small,lattice-periodic crystal images.The proposed algorithm is then validated with simulated HRTEM images of graphene and applied to the experimental images of a molybdenum disulfide(MoS_(2))monolayer with 25 variables(14 aberrations)resolved in wide ranges.Using these measured parameters,the phases of the exit-wave functions are reconstructed for each image in a focal series of MoS_(2)monolayers.The images were acquired due to the unexpected movement of the specimen holder.Four-dimensional data extraction reveals time-varying atomic structures and ripple.In particular,the atomic evolution of the sulfur-vacancy point and line defects,as well as the edge structure near the amorphous,is visualized as the resolution has been improved from about 1.75?to 0.9 A.This method can help salvage important transmission electron microscope images and is beneficial for the images obtained from electron microscopes with average stability.

基于MPC的轮毂电机驱动车辆轨迹跟踪控制

为解决轮毂电机驱动车辆在复杂工况下轨迹跟踪精度低、横摆稳定性差的问题,提出一种协调控制方法。通过神经网络动态辨识以车辆质心侧偏角-质心侧偏角速度相平面划分的稳定域边界,并基于边界线特征值得出失稳因子;将该因子作为目标函数权重影响参数,构建轨迹跟踪MPC模型,输出前轮转角和四轮力矩控制量,同时,采用滑模控制方法防止车轮过度滑转。仿真表明,该轨迹跟踪控制策略与单目标MPC轨迹跟踪控制策略相比,在低附着路面车辆最大横向跟踪误差降低了61.7%,最大质心侧偏角降低了75.7%,车辆高速行驶时,运动状态均保持在稳定域范围内,实现了轨迹跟踪精度与车辆横摆稳定性能的综合提升。

融合全局选择与局部区分的车辆重识别网络

针对跨镜头多视角差异导致车辆重识别面临的不同视角、复杂背景和光照强度等干扰问题,提出了一种融合全局选择与局部区分的车辆重识别网络。基于Resnet50骨干网络,设计了融合全局特征与局部特征的三分支互补网络,利用全局分支学习车辆的整体外观信息,局部分支捕获车辆的差异性细节信息。基于注意力机制提出了上下文特征选择模块(context feature selection module,CFSM),有效分离了车辆信息与复杂背景信息,并提出了一种细节特征增强模块(detail feature enhancement module,DFEM),利用部件之间的相对位置信息强化多粒度特征细节信息的学习。提出了一种权值自适应平衡策略,联合多损失函数进行训练。实验结果表明,所提网络在VeRi-776数据集上的mAP、CMC@1和CMC@5分别达到73.2%、93.4%和97.3%;在VehicleID数据集的大规模测试子集上,CMC@1和CMC@5分别达到75.0%和92.7%。与对比网络相比,所提网络具有较高的识别率和鲁棒性。

基于Vision Transformer的车辆重识别模型优化

针对车辆重识别任务中样本类内差异性大和类间相似度高的问题,提出了一种Vision Transformer框架下的车辆重识别方法。设计一种关键区域选择模块,整合Transformer中注意力分数矩阵,加强车辆的具有辨别性区域的关注程度,减小局部区域过度集中的注意力权重;构建一种包含对比损失和中心损失的混合损失函数,对比损失函数的引入增强了模型捕捉和比较样本之间的差异的能力,中心损失使得同一类别的样本更加紧密地聚集在一起,增强类间样本的区分度。实验结果验证了其有效性。

基于改进YOLOv7-Tiny的车辆检测研究

为了提高计算机识别检测车辆的准确度与速度,提出了一种基于改进YOLOv7-Tiny的车辆检测算法。在众多目标检测模型中,YOLOv7有着非常快的检测速度和较高的检测精度,非常适用于实时检测任务。在原YOLOv7-Tiny模型的基础上进行改进,将最浅层ELAN-T模块纳入特征金字塔,通过跳跃连接的方式将浅层特征与深层特征跨层融合,使输出的特征信息更加丰富。同时引入SE注意力机制,将计算资源分配给对当前任务更为关键的信息。并且更换了非线性激活函数HardSwish,以提高模型的表达能力。在华为发布的2D自动驾驶数据集SODA10M上进行实验,结果表明,改进后的模型对所有四种目标的检测精度都有所提高,平均精度mAP@0.5达到了66.1%,比原YOLOv7-Tiny模型61.0%提升了5.1%。

基于地铁车辆空压机出口压力空气含水量影响因素分析的正交试验设计方法

[目的]地铁列车空压机出口压力空气干燥度是一个非常重要的指标。如果压力空气中含水量高,就会影响各用风设备的性能,进而影响行车安全。上海轨道交通运维人员在汛期对某些线路列车进行检修时发现,列车供风管路中存在液态水,这是因压力空气中含水量过高而造成的。需要在现有条件下,研究如何调整维修策略,进而改善空压机在汛期的干燥能力。[方法]通过正交试验设计方法,设计了用于分析空压机出口压力空气含水量影响因素的试验方案,给出了使用田口方法分析数据进而确定主要影响因素的方法。[结果及结论]运用该试验方案进行试验,能够分析出影响列车空压机出口压力空气含水量的主要因素,运维人员可据此结论制定适用于汛期的列车空压机维修策略。该正交试验设计方法也可以推广应用至城市轨道交通其他设备运维策略的完善与优化工作之中。

光热成像技术在城市轨道交通车辆智能运维中的应用研究

[目的]目前,城市轨道交通列车在日常运行中缺少对轴箱、齿轮箱、牵引电机等关键部件运行状态的监测手段,无法实现关键部件的故障预警和跟踪,难以满足列车精准维修的要求,为适应车辆智能运维技术发展需要,需研究解决列车关键部件运行状态监测及故障预警问题。[方法]结合铁路车辆综合监测技术应用现状,提出在线路轨旁设置基于光热成像技术和声学诊断的综合监测设备,采集列车运行中关键部件的温度和噪声等信息,通过建立故障判定模型,实现对列车关键部件的动态监测和故障诊断。[结果及结论]光热成像技术具有探测部位多、探测覆盖范围大、车型适应性强、探测准确、探测结果直观等特点,与声学诊断等技术相结合,可实现列车关键部件运行状态实时监测和故障诊断,为列车实施精准维修提供技术支撑。

基于单线激光雷达的果园车辆地头导航方法

果园车辆导航包括行间引导和地头转向,地头转向主要采用开环控制,难以适应果园地面不平整以及车辆有初始偏差的复杂情况,阻碍了果园智能化作业的全程覆盖。针对上述问题,提出一种基于单线激光雷达(light detection and ranging,LiDAR)的果园车辆地头导航方法,引导车辆沿圆弧导航路径实施地头转向。导航电控系统采用上下位机结构,上位机首先读取单线LiDAR实时采集的果园数据,从中提取圆弧导航路径并计算导航偏差;然后采用纯追踪算法求取车辆目标角速度,将其发送给下位机,下位机控制车辆按照目标角速度转向。实验在株行距5 m×5 m的核桃园进行,在车速0.2 m/s、初始横向偏差±25和±50 cm、航向偏差±10°和±25°的条件下,该方法能够在地头转向结束后将横向偏差控制在±9 cm以内、航向偏差控制在±5°以内。本研究方法是对基于单线LiDAR的果园车辆导航方法的补充和完善,实验结果表明,该方法的抗干扰能力较强,能够有效削弱土地不平整和车辆初始偏差给地头导航带来的影响。

高职铁道车辆技术专业课程思政建设研究与实践--以“铁道车辆机械检修”课程为例

对专业课程实施课程思政建设是实现综合育人的重要举措。文章基于调研分析,找出课程思政建设存在的问题并借鉴先进案例,结合铁道车辆技术专业特点,深入探索课程思政建设路径,从整体设计、教学实施、评价诊断、延伸拓展等方面探讨课程思政的建设。高职院校应响应国家政策,多种举措并行,加强课程思政建设,服务高职人才综合素质的培养,使专业课程显性教育与隐性教育协同作用,实现铁道车辆技术专业高质量育人的目标。

卷取温度对铁道车辆用高耐候钢显微组织和力学性能影响

通过中试热轧试验、拉伸试验机、维氏硬度计、光学显微镜(OM)、扫描电镜(SEM)、能谱仪(EDS)研究了卷取温度对高耐候钢显微组织和力学性能的影响。研究结果表明:卷取温度显著影响高耐候钢的显微组织和力学性能。卷取温度由540℃提高到700℃,屈服强度下降189 MPa,抗拉强度下降228 MPa,延伸率提高9.5%,HV_(0.5)降低57。卷取温度为540℃和620℃时,显微组织均为准多边形铁素体和粒状贝氏体,粒状贝氏体体积分数分别为42.7%和26%。随着粒状贝氏体体积分数增加,高耐候钢的强度显著提高,延伸率大幅下降。在卷取温度为700℃时,显微组织为准多边形铁素体和沿晶界弥散分布的颗粒状碳化物,试验钢板具有优异的强塑性匹配。

履带车辆土壤承载力学特性反演

针对履带车辆下方土壤参数反演存在的问题,提出一种履带车辆土壤承载力学特性反演方法。基于履带车辆-地面耦合系统作用机理,获得履带车辆各负重轮下方最大接地压力与土壤参数的关系,以余弦函数表征履带车辆接地压力分布,探究土壤参数对接地压力的影响规律。在此基础上,基于最小二乘支持向量机对履带下方的土壤承载力学参数进行反演。结果表明,最小二乘支持向量机对基于履带车辆的土壤承载力学特性反演具有较高准确性,在重黏土和轻黏土两种情况下的最大误差略大于10%,其余情况下最大误差皆在10%之内。

VR技术在铁道机车车辆制造与维护专业教学中的应用

教育信息化背景下,VR技术在教学中的应用,不仅可赋能学生学习,更能助力教师教学。为了让VR技术在专业课教学中进一步发挥优势作用,增强学生的专业课程学习效果与体验,本文对VR技术机理、铁道机车车辆制造与维护专业定位进行了分析,结合高校铁道机车车辆制造与维护专业教学现状,以问题为导向,提出VR技术在“机车车辆车体与走行部件检修与装调”课程教学中的应用措施,以期为促进专业教学模式改革、提升课程教学水平提供参考。

轨道交通车辆车门系统生命周期环境效益评价

为了推动轨道交通车辆绿色制造的进程,以车门系统为研究对象,建立生命周期模型,构建环境效益评价指标体系,评价车门系统生命周期生产、运输、使用、废弃物处理阶段的主要环境影响,探究电力能源结构替代和焊接工艺变化对车门系统生命周期的环境影响。结果表明:车门系统生命周期不同阶段的环境影响从大到小的排序为生产阶段、废弃物处理阶段、使用阶段和运输阶段;环境影响评价指标从大到小依次为化石能源消耗潜值、温室气体排放潜值和环境酸化潜值;减少生产阶段和废弃物处理阶段的化石能源消耗和温室气体排放有助于产品生命周期环境效益提升;生产阶段中通过减少火电占比,优化电力结构,全球变暖潜值、化石能源消耗潜值和环境酸化潜值可分别降低38.96%、38.90%、38.86%;采用气体金属弧焊工艺替代保护性金属弧焊工艺,全球变暖潜值、化石能源消耗潜值和环境酸化潜值可分别降低64.71%、58.94%、59.85%。

轮毂电动机驱动车辆并联式复合制动策略

基于并联式轮毂电动机驱动实车构型,分析影响制动能量回收效率的主要因素,采用联邦卡尔曼滤波纵向车速估计算法,提出了三层式复合制动控制策略.该策略中的制动决策层根据踏板输入信号和行驶状态来识别制动工况,并进入相应的制动模式.制动控制层根据制动决策层的指令,在常规制动工况下通过粒子群优化(PSO)算法优化前、后轮电制动力矩分配,实现电池有效回收效率最大化;紧急制动工况下,通过模糊自整定PID算法实现车辆在不同附着路面上各轮滑移率的安全有效控制,优化了制动防抱死控制的鲁棒性,提高制动安全性.制动执行层响应制动控制层的指令,设计了电制动补偿机制,以迅速补偿调节各轮轮缸的压力误差,提高车辆制动稳定性.最后实车验证了该复合控制策略的有效性.

A Hybrid Cybersecurity Algorithm for Digital Image Transmission over Advanced Communication Channel Models

The efficient transmission of images,which plays a large role inwireless communication systems,poses a significant challenge in the growth of multimedia technology.High-quality images require well-tuned communication standards.The Single Carrier Frequency Division Multiple Access(SC-FDMA)is adopted for broadband wireless communications,because of its low sensitivity to carrier frequency offsets and low Peak-to-Average Power Ratio(PAPR).Data transmission through open-channel networks requires much concentration on security,reliability,and integrity.The data need a space away fromunauthorized access,modification,or deletion.These requirements are to be fulfilled by digital image watermarking and encryption.This paper ismainly concerned with secure image communication over the wireless SC-FDMA systemas an adopted communication standard.It introduces a robust image communication framework over SC-FDMA that comprises digital image watermarking and encryption to improve image security,while maintaining a high-quality reconstruction of images at the receiver side.The proposed framework allows image watermarking based on the Discrete Cosine Transform(DCT)merged with the Singular Value Decomposition(SVD)in the so-called DCT-SVD watermarking.In addition,image encryption is implemented based on chaos and DNA encoding.The encrypted watermarked images are then transmitted through the wireless SC-FDMA system.The linearMinimumMean Square Error(MMSE)equalizer is investigated in this paper to mitigate the effect of channel fading and noise on the transmitted images.Two subcarrier mapping schemes,namely localized and interleaved schemes,are compared in this paper.The study depends on different channelmodels,namely PedestrianAandVehicularA,with a modulation technique namedQuadratureAmplitude Modulation(QAM).Extensive simulation experiments are conducted and introduced in this paper for efficient transmission of encrypted watermarked images.In addition,different variants of SC-FDMA based on the Discrete Wavelet Transform(DWT),Discrete Cosine Transform(DCT),and Fast Fourier Transform(FFT)are considered and compared for the image communication task.The simulation results and comparison demonstrate clearly that DWT-SC-FDMAis better suited to the transmission of the digital images in the case of PedestrianAchannels,while the DCT-SC-FDMA is better suited to the transmission of the digital images in the case of Vehicular A channels.

Real-Time 4-Mode MDM Transmission Using Commercial 400G OTN Transceivers and All-Fiber Mode Multiplexers

Weakly-coupled mode division multiplexing(MDM)technique is considered a promising candidate to enhance the capacity of an optical transmission system,in which mode multiplexers/demultiplexers(MMUX/MDEMUX)with low insertion loss and modal crosstalk are the key components.In this paper,a low-modal-crosstalk 4-mode MMUX/MDEMUX for the weakly-coupled triple-ring-core few-mode fiber(TRC-FMF)is designed and fabricated with side-polishing processing.The measurement results show that a pair of MMUX/MDEMUX and 25 km weakly-coupled TRC-FMF MDM link achieve low modal crosstalk of lower than−17.5 dB and insertion loss of lower than 11.56 dB for all the four modes.Based on the TRC-FMF and all-fiber MMUX/MDEMUX,an experiment for 25 km real-time 4-mode 3-λwavelength division multiplexing(WDM)-MDM transmission is conducted using commercial 400G optical transport network(OTN)transceivers.The experimental results prove weakly-coupled MDM techniques facilitate a smooth upgrade of the optical transmission system.

Mother-to-Child Transmission of Human Papillomavirus in Burkina Faso

Introduction: Human papillomavirus (HPV) infection is the most widespread sexually transmitted infection in the world. Today, there is growing evidence that HPV can be transmitted early in life, and one potential route is mother-to-child transmission. Data on this route of HPV transmission are scarce in Africa and particularly in Burkina Faso, where no data on the subject are yet available. The aim of our study was to estimate the rate of mother-to-child transmission of HPV infection and to identify circulating genotypes. Methodology: Cervico-uterine samples were collected from 100 full-term pregnant women and, buccal samples were obtained from their newborns at Hopital Saint Camille de Ouagadougou (HOSCO) by the specialist physician. HPV DNA amplification and genotyping were performed by PCR followed by hybridization using the HPV Direct Flow Chips kit, detecting 36 genotypes including 18 high-risk and 18 low-risk. Results: The prevalence of HPV in newborns was 8% (8/100). Six (6) HPV-positive neonates had HPV-positive mothers, while 2 HPV-positive neonates had HPV-negative mothers. The vertical transmission rate was 26.09% (6/23). Mother-newborn genotypes were concordant. However, the genotype profile of the newborns was more restricted than that of the mothers. Conclusion: HPV DNA was found in 8% of newborns in our study. The genotype profile of the mother-newborn pair was concordant. Asymptomatic HPV infection in a pregnant woman could constitute a risk factor for vertical transmission.

Effects of cross-Kerr coupling on transmission spectrum of double-cavity optomechanical system

We theoretically study the transmission spectrum of the cavity field in a double-cavity optomechanical system with cross-Kerr(CK) effect. The system consists of two tunneling coupling optomechanical cavities with a mechanical resonator as a coupling interface. By doping CK medium into the mechanical resonator, CK couplings between the cavity fields and the mechanical resonator are introduced. We investigate the effects of CK coupling strength on the transmission spectrum of the cavity field, including the transmission rate, nonreciprocity and four-wave mixing(FWM). We find that the transmission spectrum of the probe field can show two obvious transparent windows, which can be widened by increasing the CK coupling strength. For the transmission between the two cavity fields, the perfect nonreciprocity and reciprocity are present and modulated by CK coupling and phase difference between two effective optomechanical couplings. In addition, the effects of the optomechanical and CK couplings on FWM show that the single peak of FWM is split into three symmetrical peaks due to the introduction of the CK effect.

Revealing the microstructures of metal halide perovskite thin films via advancedtransmission electron microscopy

Metal halide perovskites (MHPs) are excellent semiconductors that have led to breakthroughs in applications in thinfilmsolar cells, detectors, and light-emitting diodes due to their remarkable optoelectronic properties and defect tolerance.However, the performance and stability of MHP-based devices are significantly influenced by their microstructures includingthe formation of defects, composition fluctuations, structural inhomogeneity, etc. Transmission electron microscopy(TEM) is a powerful tool for direct observation of microstructure at the atomic-scale resolution and has been used to correlatethe microstructure and performance of MHP-based devices. In this review, we highlight the application of TEMtechniques in revealing the microstructures of MHP thin films at the atomic scale. The results provide critical understandingof the performance of MHP devices and guide the design of strategies for improving the performance and stability ofMHP devices.

Intergeneric chromosome-specific painting reveals differential chromosomal transmission from Tripidium arundinaceum in sugarcane progeny

Sugarcane has recently attracted increasing attention for its potential as a source of sugar and bioethanol,so increasing its yield is essential to ensure the sugar security and bioenergy production.Intergeneric hybridization is a highly efficient method to produce new genetic variants of crop plants,particularly those species with high ploidy such as sugarcane(Saccharum spp.).Tripidium arundinaceum exhibits many desirable agronomic traits,and has been widely studied to produce hybrids with improved stress tolerance and other characteristics in sugarcane breeding.However,the genetic relationship between T.arundinaceum and Saccharum species,and the individual T.arundinaceum chromosomal compositions in sugarcane hybrids are still elusive.Here we used whole-genome single-nucleotide polymorphisms(SNPs)to ascertain the phylogenetic relationships between these species and found that T.arundinaceum is more closely related to Saccharum than Sorghum,in contrast to the previous narrow genetic analyses using chloroplast DNA.Additionally,oligonucleotide(oligo)-based chromosome-specific painting derived from Saccharum officinarum was able to distinctly identify the chromosomes of T.arundinaceum.We developed the oligo-genomic in situ hybridization(GISH)system for the first time,to unveil the novel chromosome translocations and the transmission of individual T.arundinaceum chromosomes in sugarcane progeny.Notably,we discovered that the chromosomal transmission of T.arundinaceum exhibited several different inheritance modes,including n,2n,and over 2n in the BC1 progenies.Such inheritance patterns may have resulted from first division restitution(FDR)or FDR+nondisjunction of a chromosome with the sister chromatids in the second meiosis division/second division restitution(FDR+NSC/SDR)model during meiosis.These results will be of substantial benefit for the further selection of T.arundinaceum chromosomes for sugarcane genetic improvement.