Hot route analysis method based on density clustering

摘要

In the present invention, the target route is characterized as a route point set consisting of some route points to construct a similarity distance matrix, and the similarity between the two route point sets is compared, and then the similarity distance matrix and distance are compared. A cluster consisting of a set of root points is iteratively calculated by density clustering based on the threshold ε and the density threshold MinPts, and finally, the most frequent value of the route set of each cluster is output as a target hot route. We disclosed a method for analyzing hot routes based on clustering. The present invention provides (1) a method for comparing the similarity of target root point sets, (2) a certain degree of flexibility and robustness in selecting the density threshold MinPts, and (3). ) It has the advantage that the calculation cost is low and it is realized by an engineering method. [Selection diagram] Fig. 1