基于双目视觉的卫星太阳翼模态辨识方法

Modal Identification Method of Satellite Solar Wings Based on Binocular Vision

在解决卫星柔性太阳翼模态辨识问题时,传统接触式测量方法会对太阳翼本身动力学模型造成影响,附加测量靶标的方法,增加了额外的物理元器件和系统的复杂性,且不便于在轨实际应用。鉴于此,本文提出一种无接触、无标识点的光学测量方法。首先,用星载双目视觉系统采集柔性太阳翼振动图像,并提取太阳翼边缘角点作为特征点,测量这些特征点在时域下的位置变化情况和速度变化情况。然后,利用特征系统实现算法(ERA)来辨识太阳翼模态信息。最后,设计了物理模型试验。试验结果表明:本文提出的双目视觉测量方法装置简单,测量精度高,能够完成对太阳翼的模态辨识,模态辨识误差约为3.797%。本文所提出的基于双目相机的太阳翼模态辨识方法,为空间大挠性结构振动参数的在轨测量提供了参考。

In solving the problem of modal identification of solar wings of satellites,the traditional contact measurement method affects the dynamical model of the solar wing itself.The method of additional measurement targets adds additional physical components.This increases the complexity of the system.And the traditional method is not conducive to the practical application in orbit.Given this,an optical measurement method without contact and marking points is proposed in this paper.First,the flexible solar wing vibration images are acquired with a satellite-based binocular vision system.The corner points of the solar wing edges are extracted as feature points.The position variation and velocity variation of these feature points in the time domain are measured.Then,the Eigensystem Realization Algorithm(ERA)is used to identify the solar wing modal information.Finally,physical model tests are designed.The test results show that the binocular vision measurement method proposed in this paper has a simple device and high measurement accuracy.This method can complete the modal discrimination of the solar wings with a modal discrimination error of about 3.797%.The proposed binocular camera-based modal discrimination method for solar wing provides a reference for in-orbit measurement of vibration parameters of large flexible structures in space.