【摘要】 通过对不同空间碎片的光谱进行分析,实现对不同空间碎片的识别和分类。采用了移动平均光滑法和最大最小归一化对原始空间碎片光谱进行预处理,然后运用概率神经网络对经预处理后的空间碎片光谱进行分类。预测结果表明,移动平均平滑和归一化结合概率神经网络的空间碎片分类识别准确率达到了99.23%,说明文中提出的基于光谱技术的识别方法具有很好的分类能力。更多还原

【Abstract】 By analyzing spectra of different space debris,the recognition and classification of the different space debris are realized.In this paper,a moving average smoothing and maximum minimum normalization methods are applied to preprocess the original space debris spectra,and the probabilistic neural network is used to classify the spectra.The predicted results show that:the classification accuracy of space debris runs up to 99.23% by moving average smoothing and normalization andprobabilistic neural network.It indicates that the identification method based on spectrum technology had good classification ability.更多还原