中文说明：
一种基于粒子群K均值聚类算法的电梯交通模式识别方法。该方法通过对此前一周的原始客流数据进行聚类分析，得到相应交通模式的聚类中心坐标[2]。针对实时变化的交通流数据，采集5 mins时段客流数据，根据最近邻原则划分其归属的聚类中心，从而识别出当前的交通模式。
English Description：
An elevator traffic pattern recognition method based on particle swarm K-means clustering algorithm. Through clustering analysis of the original passenger flow data of the previous week, the cluster center coordinates of the corresponding traffic modes are obtained [2]. According to the real-time traffic flow data, the 5-minute passenger flow data is collected, and the clustering center is divided according to the nearest neighbor principle, so as to identify the current traffic mode