论文部分内容阅读
有效的甄别、掌握空间碎片的状态,是合理利用太空资源和在轨航天器规避风险的前提,为了安全、持续地开发和利用空间资源,基于散射光谱,使用夹角余弦和欧式距离嵌入K临近法(KNN-AC-ED),并与经典的朴素贝叶斯分类器作为对比,对实验室测得的空间碎片中常用的四种材质进行分类识别。分类识别的结果显示,KNN-AC-ED法比经典的朴素贝叶斯分类器总体分类精度高4%。研究表明:朴素贝叶斯分类器需要提取每个光谱曲线的三个特征,而特征提取费时费力;KNN-AC-ED法利用光谱线型和强度两种信息,不仅全面的反应了光谱的信息,且计算相对更快捷。通过对空间碎片常用材质的分类,为进一步研究空间碎片的尺寸、材质等信息提供一定的借鉴意义。
Effective screening and grasping of the state of space debris are the prerequisites for rational use of space resources and on-orbit spacecraft to avoid risks. In order to safely and continuously develop and utilize space resources, based on scattering spectra, cosine of cosine and Euclidean distance are embedded in K (KNN-AC-ED). Compared with the classical naive Bayesian classifier, the four kinds of materials commonly used in the laboratory for space debris classification are identified. The results of classification and identification show that the KNN-AC-ED method is 4% more accurate than the classical naive Bayes classifier. The research shows that the Naive Bayesian classifier needs to extract three features of each spectral curve, and the feature extraction takes time and effort. The KNN-AC-ED method uses both spectral line and intensity information to reflect not only the spectral information , And the calculation is relatively faster. Through the classification of common materials for space debris, it provides some reference for further study on the size, material and other information of space debris.