A multi-vehicle driving behavior pattern recognition method, system, device and medium

Abstract

This invention discloses a method, system, device, and medium for recognizing driving behavior patterns across multiple vehicle models, relating to the fields of intelligent transportation and autonomous driving technology. The method includes the following steps: extracting vehicle trajectory data and classifying it by vehicle model, constructing corresponding trajectory sample subsets for each model; extracting cluster centers as several prototype trajectories, and mapping each vehicle trajectory data into a fixed-dimensional morphological feature vector by obtaining the Dynamic Time Warped (DTW) distance between each vehicle trajectory data and each prototype trajectory, serving as a multi-dimensional trajectory feature set; and performing unsupervised clustering analysis on the vehicle trajectory data for each vehicle model; assigning behavior pattern labels to each vehicle trajectory data based on the clustering results, thus forming a multi-vehicle driving behavior pattern recognition result. This invention fully considers the essential differences between different vehicle models, avoiding the problem of blurred behavior pattern boundaries caused by mixing heterogeneous vehicle model data for modeling, and improving the accuracy and specificity of driving behavior pattern recognition.