A shortcut to marking 3D target curves on curved surface via a galvanometric laser scanner

Marking arbitrary three-dimensional(3D) target curves on given objects with curved surface is required in many industrial fields, such as fabric prepreg placement in composite material part fabrication, product assembly, surface painting for decoration, etc. A shortcut to the solution of this intractable problem is proposed by utilizing a galvanometric laser scanner(GLS) with the aid of a camera. Without using the existing tedious GLS calibration procedures,the proposed method directly establishes a mapping between the 3D coordinates of the laser spots on the object surface and the control voltages of the scanner. A single-hidden layer feedforward neural network(SLFN) is employed to model the mapping. By projecting a dense grid of laser spots on the object to be marked and simultaneously taking only one image, the SLFN model is trained in minutes via a linear solving mechanism. Experiments demonstrate that the trained SLFN model has a good generalization performance for marking 3D target curves. The 3D laser marking errors on experimental objects are less than 0.5 mm. The proposed method is especially suitable for on-site use and can be conveniently extended to multiple GLSs for marking large complex objects.     

Real-time object tracking via least squares transformation in spatial and Fourier domains for unmanned aerial vehicles

This paper addresses the problem of real-time object tracking for unmanned aerial vehicles. We consider the task of object tracking as a classification problem. Training a good classifier always needs a huge number of samples, which is always time-consuming and not suitable for realtime applications. In this paper, we transform the large-scale least-squares problem in the spatial domain to a series of small-scale least-squares problems with constraints in the Fourier domain using the correlation filter technique. Then, this problem is efficiently solved by two stages. In the first stage, a fast method based on recursive least squares is used to solve the correlation filter problem without constraints in the Fourier domain. In the second stage, a weight matrix is constructed to prune the solution attained in the first stage to approach the constraints in the spatial domain. Then, the pruned classifier is used for tracking. To evaluate proposed tracker's performance, comprehensive experiments are conducted on challenging aerial sequences in the UAV123 dataset. Experimental results demonstrate that proposed approach achieves a state-ofthe-art tracking performance in aerial sequences and operates at a mean speed of beyond 40 frames/s. For further analysis of proposed tracker's robustness, extensive experiments are also performed on recent benchmarks OTB50, OTB100, and VOT2016.     

磁场对高电压霍尔推力器性能影响研究

为研究磁场构型、磁场强度对高电压霍尔推力器主要性能的影响机理,以HET-200推力器为对象,采用固定磁场构型和不固定磁场构型两种方案进行实验研究。结果表明,在不固定构型的情况下增强磁场,放电电流存在两个极小值,效率也存在相应的两个极大值;但当固定磁场构型时,则放电电流只有一个最小值,效率也只有一个最大值。以最优效率下的磁场构型为基准,获得了磁场与电压、磁场与流量的匹配关系分别为B_(r, max)∝V~(0.7),B_(r, max)∝m。     

300 MVA脉冲发电机组滑差方式起动过程研究

介绍了300 MVA脉冲发电机组的滑差装置。对机组采取滑差方式的起动过程进行了分析和等效建模。根据机组实际起动过程电气波形,利用非线性规划模型,建立了滑差装置液体电阻等效仿真模型,并代入机组参数对起动过程进行仿真,仿真结果和实际波形吻合,验证了模型的正确性。在此基础上,为了使机组的转速达到额定转速,提出在滑差装置液体电阻低位时并入两级固体金属电阻后短接的方案,给出了两级固体金属电阻的阻值等理论参数,并通过仿真验证了方案的可行性。建立的仿真模型可以作为研究不同参数、不同控制策略下的机组起动过程动态特性的重要工具。     

基于改进SSD的舰船目标精细化检测方法

以SSD为代表的主流深度学习方法在目标检测领域取得了显著的成绩，但由于该类方法只能以矩形框给出目标的概略位置，检测结果具有很大的背景冗余区域，特别是港口密集停泊的舰船在图像中会出现区域重叠，导致误检和漏检。针对以上问题，提出了一种具有旋转不变性的舰船目标精细化检测方法，该方法综合利用可变形卷积、可变形池化、旋转的边框回归和旋转的非极大值抑制等模块的优点，借鉴MobileNet架构对网络加速，通过学习密集区域目标的几何形变，有效预测目标的旋转角度，最终以旋转的矩形框给出目标的位置。实验结果表明，该算法可实现多类舰船目标类型区分和目标朝向判定的功能，有效地解决了实际应用中的目标精确定位定向难题，提高了自动目标识别的精确性，并满足工程应用的实时性要求。     

Early analysis of precise orbit and clock offset determination for the satellites of the global BeiDou-3 system

Eight new-generation BeiDou satellites (BeiDou-3) have been launched into Medium Earth Orbit (MEO), allowing for global coverage since March 2018, and they are equipped with new hydrogen atomic clocks and updated rubidium clocks. Firstly, we analyzed the signals for the carrier-to-noise-density ratio (C/N0) and pseudorange multipath (MP) by using international GNSS (Global Navigation Satellite System) Monitoring and Assessment System (iGMAS) station data, and found that B1C has a lower C/N0, and B2a has the same level of C/N0 as the B1I and B3I signals. For pseudorange multipath, compared with the BeiDou-2 satellites, the obvious systematic variation of MP scatters related to the elevation angle is greatly improved for the BeiDou-3 and BeiDou-3e satellites signals. For the signals of the BeiDou-3 satellites, the order of the Root Mean Square (RMS) values of multipath and noise is B3I?<?B1I?<?B2a?<?B1C. Then, the comparison of the precise orbit determination and clock offset determination for the BeiDou-2, BeiDou-3, and BeiDou-3 experimental (BeiDou-3e) satellites was done by using 10 stations from iGMAS. The 3D precision of the 24?h orbit overlap is 24.55, 25.61, and 23.35?cm for the BeiDou-3, BeiDou-3e, and BeiDou-2 satellites, respectively. BeiDou-3 satellite has a comparable precision to that of the BeiDou-2 satellite. For the precision of clock offset estimation, the Standard Deviation (STD) of the BeiDou-3 MEO satellite is 0.350?ns, which is an improvement of 0.042?ns over that of the BeiDou-2 MEO satellite. The stabilities of the BeiDou-3 and BeiDou-3e onboard clocks are better than those of BeiDou-2 by factors of 2.84 and 1.61 at an averaging time of 1000 and 10,000?s, respectively.     

Design of a Non-cooperative Target Capture Mechanism

A passive compliant non-cooperative target capture mechanism is designed to maintain the non-cooperative target on-orbit. When the relative position between capture mechanism and satellite is confirmed,a pair of four-bar linkages lock the docking ring,which is used for connecting the satellite and the rocket. The mathematical model of capture mechanism and capture space is built by the Denavit-Hartenberg(D-H)method,and the torque of each joint is analyzed by the Lagrange dynamic equation. Besides,the capture condition and the torque of every joint under different capture conditions are analyzed by simulation in MSC. Adams. The results indicate that the mechanism can capture the non-cooperative target satellite in a wide range. During the process of capture,the passive compliant mechanism at the bottom can increase capture space,thereby reducing the difficulty and enhance stability of the capture.     

一种基于元路径拥塞模式挖掘的移动对象位置预测方法

路网上移动对象位置预测是许多位置相关服务的基础。目前移动对象位置预测方法没有充分考虑到轨迹数据中所蕴含的道路拥塞信息,而路网上的道路拥塞状态对移动对象的位置更新会产生巨大影响。提出基于元路径拥塞模式挖掘的方法(Meta-congestion-pattern mining,MCPM)。在离线挖掘阶段,从历史轨迹的频繁路径(元路径)的紧集中挖掘当地的拥塞模式,并对运动模式进行建模,其中采用基于k均值的聚类算法解决数据稀疏性问题。在线预测阶段根据挖掘的拥塞模式和运动模式依概率进行预测。最后通过理论分析和实验验证得出了算法的有效性,与相同条件下的精度预测(WhereNext,WN)方法相比,平均预测准确性提高了近20%,预测时间平均缩短了近50%。     

通信拓扑切换下的多飞行器协同拦截方法

针对通信拓扑切换条件下的多飞行器协同拦截问题，提出了一种基于扩张状态观测器的协同制导方法。建立协同制导设计模型，将协同拦截问题转换为视线稳定条件下的剩余飞行时间调节问题。为解决机动目标状态不确定的问题，将目标的状态视作扰动，设计扩张状态观测器来估计机动目标的状态，并在制导律中对目标的机动进行补偿。利用有限时间一致性理论进行一致性控制协议的设计，实现各飞行器剩余飞行时间的有限时间一致，并利用Lyapunov稳定性理论分析通信拓扑切换情况下闭环系统的有限时间稳定性，给出了系统一致收敛时间。仿真结果表明，在通信拓扑变换的情况下，设计的观测器能够有效估计目标状态，且协同制导律能够满足对剩余飞行时间的控制要求，进而实现协同拦截。     

飞秒激光系统图像调焦装置以及方法

飞秒激光直写技术在复杂三维微结构加工领域具有显著优势，而调焦是否精准直接影响了所加工结构的完整度.提出了在光路中临时置入调焦光源和物的图像调焦技术，通过调节物的位置使其成像面与激光聚焦面一致，从而通过清晰可分辨的成像状态间接反映激光聚焦状态.利用Zemax软件模拟分析了原飞秒激光光路与加入调焦光源和物的调焦光路，二者可实现相同加工物镜后工作距离与良好成像质量，证明了该方法的可行性.通过分析得到该过程的成像误差主要由成像镜头焦深(3.9 μm)引起，我们获得的理想调焦精度可达到1/2焦深以内.设计了单层高度为5 μm的二层圆柱结构，通过多次实验验证了所加工元件高度误差在1.5 μm范围以内，与理论分析一致，满足飞秒激光系统的调焦要求.