Bearings are essential components in mechanical systems,with their performance being a key factor influencing the reliability and service life of machinery.Bearing test rigs,as critical equipment for evaluating bearing performance,simulate diverse operating conditions to investigate failure mechanisms and performance characteristics under practical applications.These evaluations offer crucial insights for optimizing bearing design and operational parameters.Specialized working environments,such as heavy-load conditions,deep-sea operations,space applications,and extreme high/low temperatures,impose stringent demands on bearing performance.Consequently,performance assessment and lifespan evaluation of bearings in these conditions have become focal points in the development of advanced bearing test rigs.Research progress of bearing test rigs under heavy load conditions,deep-sea environments,space environments,and high/low-temperature environments are reviewed.The characteristics of bearing test rigs under different environmental simulation technologies are analyzed and an outlook is provided on the research trends in this field,offering a solid foundation for technological advancements in related areas.
High-aspect-ratio microgrooves have been widely applied in various fields,ranging from biology to electronics.However,uniformity is still a technical challenge in the fabrication of high-aspect-ratio microgrooves.Microgrooves with high aspect-ratio and uniformity are fabricated using a 355 nm nanosecond laser by a novel multiple-scan technique.The effects of pulse energy,defocusing amount,and scan number on the depth and width of microgrooves are studied.The results indicate that a slight increase in negative defocusing amount and pulse energy can significantly increase the aspect-ratio,but too large pulse energy will lead to a decrease in uniformity.Therefore,an optimized multiple-scan strategy is proposed to fabricate microgrooves with high aspect-ratio and uniformity by combining multiple processing and defocusing laser irradiations.Microgrooves with an average aspect-ratio as high as 6.84 are achieved,while depth and width uniformity are also significantly improved.The relative standard deviations(RSDs)of depth and width are as low as 4.8%and 4.4%,respectively.Due to the high aspect-ratio and uniformity,the microgrooves are expected to find broad application in fields,such as bio-medical science,3D miniature batteries,and superhydrophobic surface.
碳化硅外延设备是半导体产业链的重要设备之一,外延设备的控制系统直接决定了设备的优劣.针对用户对外延设备控制系统各项功能的需求尚未明确的问题,首先通过会议讨论、问卷调查、专家访谈等方式,收集整理出完整的用户需求清单并分类整合.其次运用改进A-Kano(Analytical Kano)问卷对用户需求及满意度进行权重评分.创新性地对问卷选项分值进行优化设计,有效确保了A-Kano的调查问卷所得到的结果与传统Kano模型没有原则上的偏差,提高了需求划分的相对准确性.进而利用A-Kano模型量化评估外延设备控制系统中各功能的需求层次属性,筛选出基础型需求、期望型需求和魅力型需求.根据需求清单及A-Kano需求实现优先级原则,制定了3 类典型需求的设计、实现与优化研发方案,通过架构设计、应用界面设计、算法优化、差异化需求实现等确保提供给用户稳定的基础型需求、极致的期望型需求和惊喜的魅力需求.最终经实际验证系统稳定、可靠,已生产出膜厚均匀性、掺杂浓度均匀性、表面粗糙度等关键指标符合质量标准的外延片,达到工艺要求,并且用户NPS调研满意度高.
SiC外延设备工作过程中经常会遇到信号干扰的问题,会对生产造成不便,更严重的则会造成设备损毁从而危险人身安全.为此,从工程实践的角度出发,针对控制系统一般性干扰问题和SiC外延设备特殊性干扰问题展开分析.基于目前已有的抗干扰技术,从干扰源、干扰途径入手,系统性地对碳化硅外延设备研发中出现的干扰问题进行分析解决.选取具有特点的通讯干扰、视频信号干扰、模拟信号干扰等干扰问题进行分析;通过RS485 隔离集线器、双绞屏蔽线、接地铜带、抗干扰磁环等,提升设备生产过程中的抗干扰性能;使用示波器、万用表等仪器进行检测.结果表明:采用所提方法能够明显改善信号质量,所提方案稳定、可靠,研究结果为SiC外延设备的研发与工作生产提供参考.
移动机器人的安全运行依赖于可靠的地图构建,现有的同步定位和地图构建(SLAM)方法通常使用相机或者激光雷达构建地图.然而使用单线激光雷达无法检测低于扫描平面的障碍物,使用深度相机虽然能获取三维空间障碍物信息,但由于相机视场过小等限制导致数据中可用信息过少.针对上述单一传感器SLAM方案中存在的缺点.提出一种单线激光雷达与深度相机数据融合的SLAM方法.首先对深度相机数据进行预处理,然后标定出深度相机与激光雷达传感器的外参矩阵,之后利用得到的传感器外参矩阵将深度相机点云数据转换至激光雷达坐标系进行数据融合,最后将融合后的激光雷达扫描数据输入GMapping算法构建二维占据栅格地图.通过搭建Gazebo仿真环境验证所提出的融合方法的可行性与有效性,仿真实验结果表明该方法能够有效融合相机与激光雷达信息,检测地面微小障碍物,建立更加完善的占用栅格地图.
Git分布式特性和分支管理能力适用于大型软件项目,每个分支彼此独立,互不影响.为提高碳化硅外延设备控制系统的精度、实时性、稳定性,提出一种基于Git进行配置管理的方法.介绍Git的特点、环境配置和常用操作;简单说明配置管理的概念并建立配置管理环境;介绍基于Git进行配置管理的具体实践,包括分支管理和版本管理.使用Git进行配置管理,可有效解决代码冲突、事务并发、文档冗余等问题,实现对项目的高效管理,也可保证版本的可溯性,有针对性地解决源代码混乱的问题.结果表明:应用Git的配置管理方法充分利用了 Git的优点,大大提高了碳化硅外延设备控制软件开发的效率和质量,为项目的成功提供了坚实的保障.
坩埚表面辐射率是影响坩埚性能的关键因素,直接决定了分子束外延的成膜质量,而错误的选型设计会降低坩埚性能.对分子束源炉内部复杂热场开展了机理研究和数值仿真计算,获得了坩埚表面辐射率对分子束源炉温度场的影响规律.研究发现降低坩埚表面辐射率可减少向外辐射热量,提高保温效果.采用ANSYS仿真分析源炉坩埚内部的温度场,发现坩埚埚口区域温度随着发射率的增加而降低:当发射率从0.05增加到0.95时,坩埚埚口区域温度最高降低90℃,验证了机理分析的结果.通过防氧化处理、精细抛光和表面喷涂特种涂料等处理方式降低坩埚材料表面粗糙度,能够降低坩埚表面辐射率.
生产乳胶丝的成套设备由国外公司垄断,价格昂贵.为此,简单介绍改进型JR-2型乳胶丝生产线;叙述乳胶丝生产流程的设计,并对几个关键的工艺参数进行分析;根据长期生产积累的经验,制定乳胶丝生产的工艺规程.投产实测表明,与原机型比较,JR-2型乳胶丝生产线工作参数调节更为方便,运行更稳定,能耗下降8%左右,产品合格率由88.5%提高到93.8%.国内外已有57条该生产线投产,价格仅为国外同类设备的60%左右.研究结果可为相关技术人员对挤出式乳胶丝成套装备进行改进设计提供依据,也可为乳胶丝生产企业的管理和质量控制人员提供参考.
针对当下变电站二次保护屏柜电缆接线仍采用传统人工验收方式,存在效率低、容易漏检、错检等问题,提出一种基于注意力机制的接线端子文本检测与识别方法.在文本检测阶段,针对接线端子弯曲倾斜、排列密集等问题,提出一种改进DBNet方法.使用Swin Transformer提取图像基础特征,搭建特征金字塔网络,提取并融合多尺度的图像特征,输出连接SE block,增强重要特征权重,使检测框定位更加精准.在文本识别阶段,提出一种改进CRNN方法,使用ResNet提取特征,并在残差模块中加入SE block,强化重要通道特征,进一步提升识别准确率.在检测和识别数据集上分别进行验证,结果表明:在检测数据集中,改进DBNet的精准率为95.6%,召回率为82.9%,调和平均数达到88.8%;在识别数据集中,改进CRNN方法的字符识别准确率达到87.2%.
When the insulator cleaning robot carries out cleaning operation on the insulator string, it may affect the static magnetic field environment on the transmission line, resulting in changes in the surrounding static magnetic field, and its magnetic field strength may exceed the limit. Therefore, it is necessary to carry out static magnetic field analysis of robot insulator string cleaning operation. In view of the robot cleaning the horizontal suspension insulator on the 110 kV transmission line, based on the ANSYS finite element analysis software, the magnitude and distribution of the static magnetic field intensity when the robot reaches the three positions of the insulator string were analyzed;at the same time, in order to reduce the impact of the robot on the surrounding static magnetic field environment, the manufacturing materials of the robot were optimized. The static magnetic field analysis results show that the magnetic induction intensity of the robot moving to three positions of the insulator string reaches the limit value when the transmission power of the 110 kV transmission line is 10～50 MW. The static magnetic field strength of aluminum material and copper-plated material was compared. The results show that when the robot manufacturing material is replaced with copper-plated material with relatively low permeability, the size of the surrounding magnetic field strength can be reduced, the relative variation of the magnetic field strength when the robot moves to different positions can be reduced, and the impact of the robot on the electromagnetic environment of the transmission line can be reduced.
The welding head of the high speed LED wire bonding machine frequently reciprocates, starts and stops under high acceleration conditions in the XYZ three-dimensional space, requiring the decoupling mechanism on the X/Y motion platform as its core part to be lightweight and have good static and dynamic characteristics. Taking the decoupling plate and cross slide platform of the key components of a wire bonding machine as the research object, the simplified component model was optimized by using finite element software for topology optimization; the static and dynamic analysis of the decoupling plate and the cross slide platform before and after optimization was conducted,and the mass, static displacement, and first and second order natural frequencies were calculated and compared to study the performance of the component before and after optimization. The results show that the maximum static displacement of the decoupling plate and the cross slide platform after optimization and reconstruction are respectively reduced by 40.69% and 35.29%, and the mass is respectively reduced by0.200 51 kg and 0.093 69 kg, the lightweight effect is significant; after optimization, the first order natural frequency of the decoupling plate has increased by 17.13%, and the natural frequency of the cross slide platform has decreased slightly. The results of this paper can provide a reference for further optimization of the motion platform structure of the bonding machine.
In order to improve the efficiency of manual stamping and change the existing manual stamping mode, an automatic stamping machine based on STM32 was designed. The modular design was adopted in the automatic sealing machine, and it was composed of main control module, image acquisition module, motor drive module, power supply module and three degrees of freedom mechanical structure. The main control module was controlled and processed by using STM32F103 controller, and the upper microcomputer was developed based on QT to obtain the images collected by the image acquisition module and fed them back to the GUI interface; the XY coordinates of the stamped position and the number of seals obtained from the input end were sent to the main control board, which driven the stepper motor with three degrees of freedom structure, at the same time, instructions were issued to the stepper motor to transmit the stamped documents, and the function of automatic seal was realized finally. By using the automatic stamp machine, the given number of actions in the specified position can be achieved, and the efficiency of repetitive tasks can be greatly improved.
在生产过程中,工业机器人需要在三维空间中找到位置和姿态,并以此为基础来执行移动和操作.但目前的定位系统存在发射信号模块和接收信号模块必须始终同步、接收模块时延长、定位精度差、需要采集待测目标到至少3个不同固定点的距离等问题.为此提出采用云计算的三维定位算法来解决机器人难以移动到预定位置或姿态的问题.首先通过位置测量来获得目标物体的精确位置或姿态,采用基于特征、坐标系和运动学的3种方法来实现工业机器人在三维空间中的定位.然后,结合智能线性模组、智能传感器、工业相机、物联网数据应用,在三维定位技术的基础上加入云计算高精度定位算法,设计了一款云计算的机器人三维定位软件系统并嵌入机器人.测试结果表明,系统接收信号强、识别度精准,信号发射的能量强度增强,其大小与标签距离成正比,定位准确,无须手动操作.多次型式试验并结合物联网数据分析验证了云计算的机器人三维定位软件的功能,证明所提方法有效解决了接受信号慢、定位偏离轨迹等问题,精度达到±0.5 mm,反应速度小于或等于0.02 m/s,重复定位无须编程控制,传感器能够与智能线性模组互相感应协同作业.
提出了一种基于二维压缩感知和同步混沌流密码的医学图像加密算法.首先,通过离散小波变换将明文图像转换为稀疏图像.然后通过迭代同步混沌流密码系统生成3个混沌序列,使用其中2个序列构造出测量矩阵对稀疏图像进行测量;第3个序列构造用于加密的伪随机序列.最后,通过对测量后的图像进行量化处理后再进行加密操作得到密文图像.实验结果表明,所提算法能够实现对CT图像、MRI图像、B超图像以及X光图像等医学图像的压缩和加密,在能够有效实现数据量减少的同时具有良好的加密性能.在与均匀量化函数及Sigmoid量化函数不同压缩比的比较中,采用Tanh量化函数的重构图像具有更稳定的峰值信噪比变化.同时,所提算法能抵御剪切攻击和噪声攻击,表现出较好的鲁棒性.
液罐车作为石油等危险化学品的主要运输工具,在转向行驶过程中,由于罐体内液体晃动产生的附加力和力矩,相比较于传统的运输固体货物的货车,液罐车更容易发生侧翻事故,因此,研究抑制罐内液体的晃动具有重要的意义.在传统刚性防波板的基础上研究了弹性防波板对液体晃动的抑制作用,采用OpenFOAM等软件建立了液体-弹性防波板耦合作用下的液体晃动模型.研究了在充液比为50%下,液罐车在受到斜坡阶跃加速度激励时,弹性防波板的相对长度(H/R,防波板的高度与罐体半径的比值)和厚度对液体晃动的影响.研究结果表明:弹性防波板的几何参数对液罐车的液体载荷转移量、横向晃动力和侧倾力矩均有明显的影响;弹性防波板具有很好的抑制液体晃动的作用,从而减小有液体晃动产生的作用在罐体上的附加力和力矩,进而提高液罐车的安全稳定的运行.
我国机场目前主要采用人工方式调度客舱清洁人员,效率低下,时常发生任务拖期现象.针对客舱清洁人员日常调度问题,考虑客舱清洁任务的服务时间窗、多需求同时服务等约束,以最小化总运营成本为目标包含延误成本和资源转移成本,建立混合整数规划模型.设计了一种结合贪婪算法生成初始解的自适应大规模领域搜索算法,采用多种移除算法和修复算法进行邻域搜索,并利用时间快速更新机制加速求解,用模拟退火算法接受邻域解.数值实验结果与Gurobi优化求解结果比较,表明本文算法在所有实验算例中大部分都优于或等于Gurobi求解结果,并都能在较短的时间内找到较为满意的可行解,从而验证了本文算法的求解质量和求解效率.
当360～800 Hz单相变频逆变器调频时,输出频率大范围变化会使得系统采样频率与输出频率比值可能不为整数.传统重复控制器此时无法精确跟踪输入信号,输出存在较大的静差,最终会导致系统输出的电能质量变差.为此采用了一种基于有限脉冲响应(FIR)滤波器的分数阶重复控制的控制策略,该策略既能实现分数延迟也能应用到相位超前设计上,利用分数阶相位超前解决了控制系统相位超前校正精度不高的问题.给出了控制系统的结构对其进行了理论分析,也给出了控制器的设计思路.最后,在MATLAB/Simulink工具中搭建了仿真模型,将重复控制与双分数阶重复控制两种控制方案做了对比,结果表明:双分数阶重复控制系统输出稳态误差更小,输出波形质量更好.
为解决失效模式与影响分析FMEA在汽车电控系统可靠性领域的不足问题,通过将系统工程方法思想与FMEA实践相结合的指导方法,研究了基于系统工程思维模块化的汽车电控系统可靠性分析技术,提出了一种应用于汽车电控系统FMEA分析的新方法.以满足国六排放法规的后处理柴油机颗粒捕集器DPF系统为案例进行FMEA分析,对DPF系统自顶向下进行结构逐层划分、功能逐项提取、逻辑逐条推理手段,评估DPF控制系统每条功能潜在失效模式.结果表明:对复杂抽象电控系统进行可视化结构分解及功能模块化分析,可识别出潜在的失效模式,最大程度避免汽车电控系统失效风险的发生,保障汽车电控系统可靠性与质量.采用该方法,可为汽车电控系统可靠性分析提供借鉴参考,为设计可靠性有效措施提供依据.
量化信息加工能力是推动工作系统突破自身限制,吸收先进生产力的关键.通过工作研究实验,观测到信息加工的不确定性以失真事件、矛盾行为等形式广泛存在于工作系统中.高熵环境中的失真事件普遍以必要非充分关系自发地形成耗散结构.基于耗散特征和信息熵公式,设计了量化工作系统信息加工能力的工作熵简化模型.由实践应用案例,引出了改造管理环境、施加低熵影响的工作熵管理方法.结果表明,在工作熵系统已受高熵环境影响的情况下,仍可使用高阶工作熵来考证量化过程,克服管理错觉.工作熵管理方法明确了工作系统提升和适配信息加工能力的方向,有助于全体管理人员从辩证角度重新审视工作系统和管理行为,为企业融合先进生产力创造机会.
