Traffic jam detection method based on GPS data

The invention discloses a traffic jam detection method based on GPS data. In order to effectively detect real-time traffic state in a certain area, the method divides a study area into clusters of different sizes for respective studying. For traffic state detection, instantaneity is important. The method carries out sampling based on sample size within clusters to reduce the amount of computing. According to the method, in order to ensure the rationality of data, acquired GPS data are pre-processed to get rid of abnormal data; on the basis of the integrity of acquired traffic information, the time-space characteristic of the GPS data is used, and a k-means algorithm is used to carry out clustering analysis on the GPS data to realize study area division in different clusters; and based on sampling based on sample size within clusters, real-time traffic state detection is carried out, which effectively improves the instantaneity. 本发明公开了种基于GPS数据的交通拥堵检测方法，本方法为了对某区域内实时交通状态进行有效的检测，将研究区域分成不同大小的簇分别进行研究；对于交通状态的检测，重要的是