基于SOM的我国民用机场分类研究

摘要

科学准确地对民用机场分类可以将机场自身定位，为机场发展及评估提供可靠依据，具有重要的现实意义。本文选用了15个影响我国机场分类的关键因素，运用神经网络技术中的自组织映射（Self-Organizing Map，文中简称SOM）将我国民用机场进行了细分，并分析了各类机场的特点。结果表明，SOM能很好地应用于我国民用机场分类，是一种科学、有效的分类方法。%Scientific and accurate classification of airports can determine their positions and can provide a reliable basis for airports’ development and assessment. 15 factors affecting the classification of our country airports were selected and the neural network technology (self-organizing mapping) was used to divide the airports into three big categories, then each airport’s features was analyzed .This results show that the SOM is properly used for airport classification in China, and it is a scientific and effective method.