摘要
由于火箭发动机极性测试依赖人工同时判断记录多个喷管的摆动方向,属于“耳听手摸”测试项目,对于摆幅小、摆动速度快的工况,人眼也难以辨识跟踪,测试有效性存在差距。从系统架构设计与应用实践的角度对测试系统进行阐述,通过加装工业摄像头并设计一套基于计算机视觉方法的测量系统,可以有效判断出各喷管摆动轨迹与极性,提高测试效率和测试覆盖性,使测试过程可追溯,保证测量的可靠性,减少人工参与环节,从而离型号测试的无人值守目标更近了一步。本系统目前已在型号中成功搭载,为人工智能技术在型号中的应用奠定了基础。
Launch vehicle engine polarity test relies on technician to simultaneously judge the swing direction of multiple nozzles which thereby is both time-consuming and low-automation.For the case of small swing and fast swing speed,it is difficult for the human eye to identify and track which results in low test effectiveness.This paper describes the test system from the perspective of system architecture design and application practice.By installing a set of cameras and designing a measurement system based on computer vision method,the oscillation trajectory and polarity of each nozzle can be effectively judged.We show that our system is able to improve test efficiency as well as test coverage,make testing process traceable and ensure measurement reliability.The system has been successfully carried out in the launch vehicle test link,which lays a foundation for the application of artificial intelligence technology in the successful taking-off of the launch vehicle.
作者
王冠
易航
张恒
廖友萍
Wang Guan;Yi Hang;Zhang Heng;Liao You-ping(Beijing Institute of Astronautical Systems Engineering,Beijing,100076)
出处
《导弹与航天运载技术》
CSCD
北大核心
2019年第4期108-113,共6页
Missiles and Space Vehicles
