中文说明：基于统计直方图结合概率神经网络，实现了对作用面结构类型的分类预测。从预测结果来看，统计直方图提取出的特征，具有很好的区分能力，而且可以通过调节划分的区间个数和节点的选取方式，达到对作用面结构的不同粒度的描述，以适用于不同目的的研究，这可能对与结构有关问题的研究具有启发性。
English Description：
Based on histogram combined with probabilistic neural networks, classification and prediction of structural type is implemented. From the forecast results, extracted Histogram feature has a very good ability to differentiate, but you can adjust the interval number and node of the Division of selection, to descriptions of the surface structure of multi-granularity, suitable for different research purposes, this may the study of structure-related issues is instructive.