基于OTSU二值化的航天器在轨环境预测方法

Prediction method of spacecraft on-orbit environment based on OTSU binarization

摘要:

以往采用统计能量法和虚拟模态综合法预测在轨环境，存在预测精准度低的问题，为了避免传统方法带来的弊端，提出了基于OTSU二值化的航天器在轨环境预测方法。使用单目视觉光学测量法测量航天器所在位置，设计三圆靶，采用亚像素处理方法获取航天器像素版图。推导出像素灰度值以及域灰度值出现频数，引入萤火虫算法进行OTSU二值化图像分割。采用三角形网格模型，以航天器在轨特性为基准生成一个二值化子图，调用宽度优先搜索算法获取最初近似最短路径。取出子图中的边，再进行细分处理，生成子图点集，将子集中的各点连接成线，得到像素点间最短路径，实现航天器在轨环境预测。由结构动力学实验结果可知，该方法所捕获的冲击响应谱与实际获取的冲击响应谱最大误差为100g，而传统方法最大误差为400 g，由此说明基于OTSU二值化方法预测精准度较高，为航天器实际运行条件制定提供了参考。

Abstract:

In the past, the statistical energy method and virtual modal synthesis method were used to predict the on-orbit environment, and there was a problem of low prediction accuracy. In order to avoid the disadvantages of the traditional method, a method for predicting the on-orbit environment of a spacecraft based on OTSU binarization was proposed. . The monocular vision optical measurement method is used to measure the position of the spacecraft, a three-round target is designed, and the subpixel processing method is used to obtain the spacecraft pixel layout. The frequency of occurrence of pixel gray value and domain gray value was derived, and the firefly algorithm was introduced to perform OTSU binary image segmentation. A triangular mesh model is used to generate a binary subgraph based on the orbital characteristics of the spacecraft. The width-first search algorithm is called to obtain the initial approximate shortest path. Take the edges in the sub-graph and subdivide them to generate a sub-graph point set. Connect the points in the sub-set into a line to get the shortest path between the pixels. From the results of structural dynamics experiments, it can be seen that the maximum error between the shock response spectrum captured by this method and the actual shock response spectrum obtained is 100 g, while the maximum error of the traditional method is 400 g. High, which provides a reference for the establishment of the actual operating conditions of the spacecraft.