| 77 | 1 | 21 |
| 下载次数 | 被引频次 | 阅读次数 |
Risley棱镜视轴指向控制技术已被应用于军事领域的飞行目标跟踪，而飞行目标时间步坐标直接影响Risley棱镜跟踪精度。因此，开展飞行目标轨迹预测将为Risley棱镜视轴指向控制提供革新规划与控制思路。结合Social-STGCNN深度学习方法对Risley棱镜控制算法进行优化。首先定义数据集获取飞行目标信息，利用STGCNN预处理，然后运用旁轴法求得Risley棱镜方位与俯仰角并进行仿真分析实验。结果表明，在对Risley棱镜电机与视轴实现精密调控的同时，取得了0.151 mm的轴间误差，在MAE指标中比单纯采用图像误差反馈的控制算法低88.29%，有效验证了本文所控制策略的优化效果与跟踪性能。
Abstract：Risley prism boresight pointing control technology has been applied to flying target tracking in the military field. The coordinates of flying target time step directly affect the tracking accuracy of Risley prism.Therefore, carrying out flight target trajectory prediction will provide innovative planning and control ideas for Risley prism boresight pointing control. This article combines the Social-STGCNN deep learning method to optimize the Risley prism control algorithm. First, the data set is defined to obtain flight target information, and STGCNN is used to preprocess. Then the paraxial method is applied to obtain the Risley prism azimuth and pitch angle. Simulation analysis experiments are conducted. The results show that while achieving precise control of the Risley prism motor and the visual axis, an inter-axis error of 0.151 mm is obtained. This is 88.29%lower in the MAE index than the control algorithm that simply uses image error feedback, which effectively verifies the optimization effect and tracking performance of the control strategy proposed.
[1]洪华杰，周远，陶忠，等. Risley棱镜在光学侦察中的应用[J].应用光学，2014,35(2):179-187.
[2]梁卫清，魏志强，袁红伟，等.小型高性能无人机载光电吊舱的发展现状与方向[J].电视技术，2022,46(7):65-68.
[3]周书芃.消色差旋转双棱镜光束指向控制技术[D].北京：中国科学院研究生院（光电技术研究所），2016.
[4]BOISSET G C,ROBERTSON B,HINTON H S. Design and constructon of an active alignment demonstrator for a free-pace optical interconnect[J]. IEEE Photonics Technology Letters,1995,7(6):676-678.
[5]LI Y. Third-order theory of the Risley-prism-based beam steering system[J]. APPLIED OPTICS,2011,50(5):679-686.
[6]LI A,GAO X,SUN W,et al. Inverse solutions for a Risley prismscanner with iterative refinement by a forward solution[J].APPLIED OPTICS,2015,54(33):9981-9989.
[7]JING L,WEIWEI R,LINHUI L,et al. PTP-STGCN:Pedestrian Trajectory Prediction Based on a Spatio-temporal Graph Convolutional Neural Network[J]. Applied Intelligence,2022,53(3):2862-2878.
[8]陈辉.基于深度图卷积网络的半监督节点分类研究[D].北京：华北电力大学，2023.
[9]ALAJLOUNI S E. Solution to the Control Problem of Laser Path Tracking Using Risley Prisms[J]. IEEE-ASME TRANSACTIONS ON MECHATRONICS,2016,21(4):1892-1899.
[10]GOVINDA C,MOKLEIV H N,AAS S R,et al. Introducing the ETH/PRIO Civil Conflict Ceasefire Dataset[J]. Journal of Conflict Resolution,2023,67(7/8):1430-1451.
[11]孙科，鄢府，范勇强，等.基于时空图联合关系路径的行人轨迹预测框架[J].无线电工程，2023,53(2):281-289.
[12]HODSON, TIMOTHY O. Root-mean-square error(RMSE)Geoscientific Model Development,2022,15(14):5481-5487.
[13]林嘉豪，章宗长，姜冲，等.基于生成对抗网络的模仿学习综述[J].计算机学报，2020,43(2):326-351.
[14]曹健，陈怡梅，李海生，等.基于图神经网络的行人轨迹预测研究综述[J].计算机工程与科学，2023,45(6):1040-1053.
[15]KUMARESAN P S,TAN K C,et al. An Efficient Stacked-LSTM Based User Clustering for 5G NOMA Systems[J].Computers, Materials&Continua,2022,72(3):6119-6140.
[16]JINSOO C,TAEHYUN O. Joint Video Super-Resolution and Frame Interpolation via Permutation Invariance[J]. Sensors,2023,23(5):2529-2529.
[17]徐胜军，杨华，李明海，等.基于双频域特征聚合的低照度图像增强[J].光电工程，2023,50(12):32-49.
[18]张文博，吴圣雨，陶冶.一种采用图像误差反馈的Risley棱镜光电跟踪系统的闭环控制方法[J].机械，2023,50(8):8-15.
基本信息:
DOI：
中图分类号:TP18;E91
引用信息:
[1]陶冶,吴圣雨,吴小龑等.基于Social-STGCNN轨迹预测的Risley棱镜跟踪控制方法[J].机械,2024,51(03):11-18+37.
基金信息:
国家重点研发计划（2022YFF0712903-1）; 中国工程物理研究院院长基金（YZJJZQ2022001）