中文摘要：

人工神经网络理论是较新的数学分支学科，神经网络分类和排序是刚引入植物生态学的分析方法。本研究应用这2种方法研究了五台山亚高山高寒草甸。结果SOFM聚类将78个样方分为8个植物群落类型，基本上代表了本地区高寒草甸的群丛类型，具有明确的生态意义；SOFM排序反映了明显的生态梯度，表明海拔是影响草甸植被生长和分布的最主要因子，坡向和坡度也有一定的作用；SOFM聚类和排序方法分析应用效果好，2种方法结合使用更好；五台山草甸需要进一步加强保护。

英文摘要：

Artificial n recently been introdu cold meadows in the eural network theory and ced to plant ecology. This Wutai Mountains. SOFM ordination are relatively work applied these two new branches of mathematics that have methods to study the subalpine high and clustering classified 78 quadrats into 8 community types, basically representing the associations of the high and cold meadows in the Wutai Mountains. This classification is meaningful in ecology. The SOFM ordination clearly reflected ecological gradients, indicating that altitude is the most important factor affecting the growth and distribution of meadow vegetation, while slope and aspect also have certain roles. SOFM clustering and ordination methods performed well in this application, and this study showed that the combination of these two methods is good for ecological analysis. The conservation ofmeadows in Wutai Mountains needs to be strengthened further.