A method of surrounding vehicle behavior recognition

Abstract

The invention discloses a surrounding vehicle behavior recognition method based on V2V communication and HMM‑GBDT hybrid model, which belongs to the field of vehicle intelligent driving, including: a. Offline training link: summarizing and dividing typical surrounding vehicle behaviors, aiming at each typical behavior Based on the real vehicle platform, the driving information of the surrounding vehicles in the real traffic scene is collected, the trajectory feature data is extracted, and the parameters of the HMM‑GBDT hybrid model are learned. b. Online detection link: the tracked target vehicle transmits the collected driving information to the main vehicle in real time, and the main vehicle combines the trajectory feature data of the two vehicles to form a new feature observation sequence, and uses the trained HMM-GBDT hybrid model to identify the target vehicle. Behavioral patterns to which vehicles belong are tracked. The present invention acquires the characteristics of the vehicle's historical trajectory by passively receiving information, avoiding the influence of traffic conditions and environmental factors on active detection; does not rely on the fixed base station in the general Internet of Vehicles system, ensures instant information transmission, and accurately recognizes the behavior of the target vehicle .

Description

technical field [0001] The invention belongs to the field of vehicle intelligent driving, and in particular relates to a surrounding vehicle behavior recognition method based on a V2V communication and HMM-GBDT hybrid model. Background technique [0002] In recent years, vehicle behavior recognition by analyzing the characteristics of vehicle historical trajectories has become one of the hot issues that researchers pay attention to. The key to behavior recognition is to learn the behavior pattern of the vehicle, establish a behavior recognition model, and then perform vehicle behavior recognition through the trained vehicle behavior recognition model, and even predict vehicle behavior. [0003] In order to provide training samples for the surrounding vehicle behavior recognition model, it is necessary to use V2V technology to exchange each other's state information to extract vehicle trajectory feature data. V2V technology was first proposed by some American car companies. ...

AI Technical Summary

Problems solved by technology

V2V must also overcome the complex communication environment. The Doppler effect caused by high-speed movement and the complex communication environment will cause serious wireless loss, and the mutual interference of communication between vehicles will also play an important role.

Images