Clustering method for urban driving track and driving behavior of vehicle

Abstract

The invention discloses a clustering method for urban driving tracks and driving behaviors of vehicles. The clustering method comprises the following steps: step 1, simplifying the tracks and then extracting corner points; step 2, carrying out clustering processing on the corner points, and constructing a spatial-temporal index; 3, compiling a track sample by using the index obtained in the step 2 to obtain a track code; 4, defining distance parameters of track coding; and step 5, carrying out clustering analysis on the trajectory under different dimensions. According to the method, the trajectory is simplified in practicability according to urban road characteristics, so that the calculation efficiency is greatly improved, and meanwhile, the traditional length trajectory data analysis problem is reasonably decomposed, so that more trajectory characteristics are obtained. According to the method, an editing distance algorithm is optimized in calculation, so that distance calculation in a specific scene is more accurate.

Description

technical field [0001] The invention relates to the technical field of data mining, in particular to a clustering method of vehicle urban driving trajectories and driving behaviors. Background technique [0002] With the popularization of in-vehicle positioning terminals and the perfect installation of collection equipment on the road, it becomes possible to manage different types of vehicles in the city and refine the optimization methods for traffic. The trajectory data of vehicle positioning is the spatial point of attachment time and motion characteristics, and the trajectory data of road collection equipment is the fixed spatial point of attachment time and graphic features. Combining the two kinds of information can basically describe all the states of the vehicle, which can be analyzed and speculated. A wealth of behavioral and personnel information is available. [0003] The mining of vehicle trajectories helps to classify urban traffic, so as to induce and manage a...

AI Technical Summary

Problems solved by technology

[0004] In the traditional clustering method for trajectory data, more attention is paid to the spatial distribution characteristics of the start and end points of the entire trajectory, where the start and end points are simply divided by day or single travel as a unit, which cannot reasonably highlight the characteristics of a certain behavior. Features; the spatial distribution features of interest are mostly measured by Euclidean distance and because of the massive amount of trajectories, the amount of calculation is huge and time and motion characteristics are not considered, so it is difficult to efficiently extract driving features

Images