摘要
在运载火箭开展射前综合测试时,控制系统在预先指定的飞行时序向末端发送一系列指令,指示末端设备做出相应动作,以达成对应的目标。这些动作主要依赖火工品完成,这些动作变化会体现在火工品等效器指示灯上。当前控制系统无法检测到末端等效器的电流情况变化,因此无法对末端动作进行有效监测,测试缺乏可靠性、准确性与可追溯性的问题亟需改善。通过综合运用人工智能、数字图像处理与计算机视觉技术,经由特征匹配、指示灯定位、指示灯状态建模等步骤,可对末端火工品等效器状态变化实现精准监测,进而实现飞行时序自动判读。与标准判据比对后,本文提出的火箭飞行时序判读系统准确率可达98.89%。
When conducting the pre-lunch comprehensive test of rockets,the control system sends a series of instructions of of the pre-designated flight sequence to the terminal equipments at the end to make corresponding actions.These actions are mainly performed by the Electric Explosive Device(EED),and these changes will be reflected in the EED Equifier indicator.Currently the control system can not detect the changes of current of the terminal equivalent,so the terminal actions can not be monitored effectively.The weakness in reliability,accuracy and traceability needs to be improved urgently.By using the comprehensive application of artificial intelligence,digital image processing and computer vision technology,through the steps of feature matching,indicator light state modeling and so on,the state changes of the explosive device equivalent can be accurately monitored,and then we could achieve the flight timing sequence automatic interpretation.Compared with the standard criterion,the accuracy of the rocket flight timing sequence interpretation system achieves 98.89%.
作者
王潇宇
王伟
刘巧珍
徐利杰
WANG Xiaoyu;WANG Wei;LIU Qiaozhen;XU Lijie(Beijing Institute of Aerospace Systems Engineering,Beijing 100076,China)
出处
《宇航总体技术》
2021年第4期43-51,共9页
Astronautical Systems Engineering Technology
基金
中国运载火箭研究院技术创新基金项目
关键词
人工智能
图像识别
飞行测试
时序判读
Artificial intelligence
Image recognition
Flight test
Timing interpretation
