The integrated monitoring system for rail transit is a highly complex control and supervision system used for the regulation and control of various facilities and services in the rail transit network. Due to its high requirement for reliability, it is necessary to establish both a primary and a backup system. When the primary system encounters issues, the backup system takes over the line control authority. When the primary system recovers, the backup system returns the line control authority. However, existing switching methods suffer from issues such as poor reliability and low stability, seriously affecting the safe operation of rail transit. We propose a switching method for the integrated monitoring system and the main and backup control centers of rail transit, aiming to improve the reliability and stability of the integrated monitoring system. Firstly, we establish primary and backup monitoring systems with consistent software and hardware, which operate independently and simultaneously, providing the basic conditions for reliable and stable switching. Secondly, we construct a line control authority switching method that effectively enhances the timeliness of line monitoring during primary system failures through automatic switching of line control authority. Finally, we demonstrate the advantages of our proposed method in terms of stability and reliability through a case study based on real scenarios.
In the wind power industry,the algorithm library plays an important role in the development of control systems as a carrier for the application and accumulation of industry knowledge.In order to solve the problem that the wind power algorithm library of foreign brand controllers is not open source and realize the independent development of wind power al-gorithm library,this paper analyzes the specific needs of the wind power industry for the algorithm library,combines the commonly used control system program design standards in the industrial control industry,proposes a design method of wind power algorithm library from the perspective of functional design and interface design,and describes the process of realizing it through IEC61131-3 language and C language.The algorithm library developed by this method has been ap-plied to large-scale wind turbines.
In order to solve the problem of partial loss of data information and structure of 3D point cloud model due to subjective and objective factors such as occlusion and noise, a partial deletion reconstruction method of 3D point cloud model based on PF-Net was proposed, and a model deletion reconstruction system was developed. Based on the deep learning PF-Net network architecture, the Batch Normal layer and the Dropout layer are introduced to normalize the original datasets in batches, which further improves the reconstruction efficiency and accuracy of some missing point cloud models. In this paper, 11 point cloud models are selected to carry out reconstruction experiments with some missing data information and structural features, and the experiments show that the proposed method has higher reconstruction efficiency than the L-Gan and PCN methods when the same dataset is used for training and testing. In the eleven test categories, the average improvement of the refactoring method in this paper is 12%-27%. The proposed method has a significant effect in dealing with the partial deletion reconstruction of small-scale models, and at the same time improves the efficiency and accuracy of reconstruction, which has good application value.
For the modeling of the single coil structure commonly used in wireless charging, Maxwell static magnetic field was used to analyze the influence of changing the longitudinal distance of the coil on the transmission characteristics of the three single coil structures when other parameters were unchanged. By comparing the performance of the three single-coil structures, a new transmitting coil structure was proposed on this basis. The magnetic field simulation showed that the new transmitting coil could generate greater magnetic field intensity and more significant energy transmission effect than the traditional coil.
Aiming at the classical knapsack problem in combinatorial optimization, in order to improve the local search ability and global search ability of the basic fireworks algorithm, an improved fireworks algorithm is proposed by combining the basic fireworks algorithm, greedy optimization strategy and simulated annealing algorithm. In order to ensure the diversity of the initial population, the Tent mapping is proposed to initialize the population ; greedy repair operator and greedy optimization operator are introduced to correct the intermediate solution ; at the same time, the simulated annealing mechanism is introduced to make the poor solution have a certain probability to be accepted to improve the ability of the algorithm to jump out of the local optimum. By solving the typical test function, it is found that the improved fireworks algorithm can accurately solve the theoretical optimal solution of the Griewank function. Compared with the basic fireworks algorithm, simulated annealing algorithm and particle swarm optimization algorithm, the improved fireworks algorithm can find the optimal value of Sphere function with higher accuracy. By solving six groups of knapsack problems with different dimensions, it is found that the improved fireworks algorithm can hit the optimal solution with 100 % probability for most test data, and hit the optimal solution with more than 70 % probability for its test data, and the standard deviation of hit times is less than 13. The experimental results show that the improved fireworks algorithm has higher solving accuracy and faster solving speed, and can effectively solve the 0-1 knapsack problem.
为了高效利用多视图数据之间的一致性和互补性信息,提出了一种多视图解耦的变分自编码器(MVDVAE,Mul-ti-View Disentangled Variational Auto-Encoder)模型.该模型针对一致性信息提出了一种基于变分自编码器(VAE,Variational Auto-Encoder)的分布对齐和加权融合策略,可以达到视图间共有信息的一致性;其次,为了保留各视图的特有信息以及解耦一致性和互补性,提出了强化重建损失,去保留采样后的互补性信息.实验结果表明,该模型相较于其他方法在三个真实数据集上都有较大的提升.
作物真菌性气传病害的特点意味着早期预警和干预能有效提高防治效果.现有孢子监测装置体积大、成本高、空间布局灵活性差,降低了病情评估的准确度.提出了一种基于无透镜成像的便携式孢子监测物联网系统.将空气中的孢子富集于载玻片上,采集 24.396 mm2 的无透镜成像,所设计的算法获得高质量重建结果,提取包括形态、幅相在内的多维度特征,实现孢子识别和分类.采用该方法对稻瘟病孢子进行富集和浓度检测,结果表明,该方法成像视场大、准确度高、检出限高,能在病害早期预警.其便携式、自动化特点有望为明确气传病害演化规律和预防流行传播提供数据支持.
针对弹药装配行业危险系数高不便于人工操作的现状,设计了一款自动输送弹体给下一装药单元的送料控制系统,该系统采用X、Y、Z直角坐标系方案,X和Y方向采用两个伺服电机实现精确定位每一颗弹体的位置,Z方向采用气缸升降和气爪完成抓取和放料动作.为节约成本和提高国产化率,该系统采用国产英威腾伺服驱动器.应用结果表明,该系统能够实现高效连续地自动化抓料和放料任务,提高了弹药装配送料环节的安全性.
为了实现了离散型企业数据的实时处理和统一化处理,有效解决产品种类多、数据标签复杂、人员检测疏漏等生产问题,在提出大数据智能管理体系的同时,搭建了全流程、一体化的"多码合一"大数据应用平台,建立了企业检测数据的实时采集、处理、分析、反馈和产线实时调整的闭环管理系统.平台突破传统产线的技术壁垒,采用固定式和运动式一体化检测系统,实现产线质量的实时监控分析,提供柔性化的技术解决方案.大数据平台融合连接MES系统、产品数据库、工厂大数据中心,通过统一的传输格式实现数据的交汇互通,有效实现制造业工厂数据的快速收集、海量存储、深度挖掘和实时分析,促进企业的转型升级.
星载SAR具有利用稳定重访周期来获取高分辨图像的优势,针对其获得的长时间时序图像,利用变化检测技术可以对固定区域中的变化信息进行提取.对于环境监测、灾害损失评估、生产能力评估有着重要意义.现有方法多针对较大尺寸的区域,面向目标级的变化检测方法较少.因此提出了一种基于Log-Ratio算子的星载SAR时序图像序列变化检测方法,以实现获取目标级尺寸的时序变化信息.该方法首先运用Log-Ratio算子将图像转换为对数图像,然后以图像序列中的一景为参考图像,将当前图像与参考图像作差获得变化图像,最后在变化图像上运用CFAR算法完成变化检测.所提算法利用Sentinel-1 SAR图像数据集进行了方法验正.
针对系统级封装SIP(System In a Package)测试需求增长,介绍了一种基于特定裸芯片的硬件测试系统设计方案.选用两种主控裸芯为主控芯片,配合硬件电路,通过软件控制,该硬件测试系统可实现程序下载和外部接口验证.该硬件测试系统配有多路串行通信接口和普通IO接口,可模拟后续待测SIP芯片引出功能.该硬件测试系统经过长时间常温环境连续工作验证,所设计的电路板工作稳定.硬件测试系统正常工作时功耗为 2.5 W,可实现RS422 串口和SPI数据传输,实现程序下载复位重载功能.该硬件测试系统为下一步待测SIP芯片的设计生产提供了参考.
在商业大楼和高层写字楼里,基于PLC对电梯进行控制能够实现安全稳定运行.但如何合理调度多部电梯,使得电梯高效运行成为电梯控制研究的热点.基于PLC控制的六部十层电梯系统,研究了电梯的群控算法.首先以模块化方案对单部电梯进行控制,在所建立的单部电梯的控制模块基础上,提出了一种基于最短距离的动态响应电梯群控调度算法.通过电梯仿真系统的验证,结果显示该算法可以实现电梯的稳定高效运行,且能够在多种情景模式下高效运行.
针对传统ADS-B用于对新型军民航空器的监视存在的时空局限性问题,提出了星基ADS-B系统.该系统首先利用合理的卫星构架,让目标被 2 个以上卫星覆盖,通过卫星内置GNSS实现时间同步,对接收的信号实时译码得到ADS-B报文并传入地面数据中心,解决了监视空域覆盖范围不足的问题;其次,地面数据中心进行位置解析和航迹生成的过程,不同的卫星下传同一条报文开展TDOA计算得到目标位置信息,然后进行比较确认有效信息,解决了ADS-B监视系统可能得到虚假目标的问题.实验结果表明,在复杂的连续飞行航班及其目标信息确认高要求的情况下,结合卫星构架,利用信号TDOA计算,在一定的误差范围内,对星基ADS-B目标真实性和有效性验证是有效的.
FINS Protocol can be used in the OMRON networks communication such as CompoBus/D,Controller Link,Ethernet.Principles of FINS protocol were introduced briefly.The serial communication method between master computer and PLC based on FINS protocol was investigated.OMRON CP1H PLC was chosen as a remote slave.The program in master computer used Visual C++ 6.0 was introduced and message exchange was realized by sending FINS command to PLC.This method can be further developed as a simple,remote and real time supervision method based on FINS protocol.
针对电子产品设计中主控制器核心硬件串口资源缺乏的问题,提出了一种基于WK2124 构建SPI接口四通道异步串口扩展器的设计方案,利用主控制器的硬件 4 线SPI资源或常规的 4 个I/O引脚即可得到 4 路增强型异步串口,该方案相较于常规的软件模拟串口时序或线路分时复用等方法更具实用性,扩展的子串口具备片上FIFO收发缓存单元,数据帧格式和波特率可单独调整,适用于串口服务器、无线DTU设备、车载设备、抄表系统和嵌入式系统中需要串口扩展的应用场景.
针对车间的设备种类和数量众多,导致底层设备的通信接口和协议的多样化且不兼容等问题,设计了一种基于OPC(OLE for Process Control)协议的边缘服务器采集软件.使用C#语言的WPF 框架、TreeJS技术,以及OPC UA(OPC Unified Architecture)标准协议,结合了边缘服务器延迟性低的优点,设计了具有用户登录、数据采集、管理、查询,3D展示等功能的上位机软件.为验证系统的可行性,对其进行测试,实现了数据的采集和Web端的 3D展示,证明该系统可以满足车间的基本需求.
针对国六排放标准对汽车发动机失火故障诊断提出的新要求,提出一种基于KPCA的发动机多传感器数据驱动的失火故障诊断模型.该方案选取了发动机传感器网络中 6 个与失火关联度较高的传感器数据,分别进行数据处理和特征提取;然后使用KPCA对特征矩阵进行降维融合,挖掘特征之间的非线性关系,从而形成对发动机运行状态相对综合的评价;再以T2 统计量和Q统计量作为指标进行失火故障诊断;最后通过发动机仿真数据集验证了该方案进行失火故障诊断的准确率.
深度学习的成功依赖于海量的训练数据,然而获取大规模有标注的数据并不容易,成本昂贵且耗时;同时由于数据在不同场景下的分布有所不同,利用某一特定场景的数据集所训练出的模型往往在其他场景表现不佳.迁移学习作为一种将知识从一个领域转移到另一个领域的方法,可以解决上述问题.深度迁移学习则是在深度学习框架下实现迁移学习的方法.提出一种基于伪标签的深度迁移学习算法,该算法以ResNet-50 为骨干,通过一种兼顾置信度和类别平衡的样本筛选机制为目标域样本提供伪标签,然后进行自训练,最终实现对目标域样本准确分类,在Office-31 数据集上的三组迁移学习任务中,平均准确率较传统算法提升 5.0%.该算法没有引入任何额外网络参数,且注重源域数据隐私,可移植性强,具有一定的实用价值.
现有的深度压缩感知重建算法在低采样率下,由于使用像素损失指导优化的网络会使得重建的图像无法有效地提取出原始图像的纹理细节,导致重建图像视觉观感较差.针对上述问题,提出了基于感知生成对抗网络的图像压缩感知重建算法,用感知损失代替像素损失,使得重建图像细节和纹理特征保留.通过对比实验表明,提出的基于感知生成对抗网络的图像压缩感知重建算法在低采样率下重建出的图像具有更强的视觉效果和真实性.
在网络中,由于配置错误或恶意路由宣告而导致的BGP前缀劫持事件,给当今的互联网带来了极大的麻烦.外包缓解措施是最近提出的一种前缀劫持缓解修复的方法.它通过将把数据同步出来传送给路由器,再提取ROA当中的信息,更新证书后,替换Update报文中的内容再进行传递的机制方法,来缓解源AS号劫持事件,使机制重定向位置以此吸引大量网络流量.吸引和重定向行为的AS号对于提高修复能力非常有效.因此,如何衡量不同AS号的缓解效果并有效选择缓解因子是外包缓解措施的关键问题.为了从网络整体来均衡能耗,延长网络生存时间,采取了Cluster路由算法改进的方法,对不同的簇树中节点关系的选择,由不同节点的能量来避免路由发生回路和绕路等现象,从而更改Cluster-list id进行严格的路由选择,实现最优路由产生.
