A driving risk identification method based on track prediction

Abstract

The invention discloses a driving risk identification method based on track prediction, and the method comprises the following steps: S1, obtaining the GPS information in real time, and generating a reference track; S2, according to the reference trajectory and GPS information, in combination with a vehicle dynamics model and a control constraint condition, generating a feasible trajectory set which is dynamic and is connected with the initial end of the reference trajectory and the terminal and a speed corresponding to each feasible trajectory in the feasible trajectory set; S3, optimizing the feasible track and the corresponding speed; S4, calculating the cost values of the feasible tracks by using a cost function according to the optimized feasible tracks and speed in combination with avehicle dynamics model and a set constraint condition, and forming a probability distribution diagram of the own vehicle driving on each feasible track, thereby drawing a feasible track distributiondiagram; Wherein the cost function is obtained according to a minimum action principle and an equivalent force method; And S5, according to the feasible track distribution map, calculating a relativeequivalent force by using the vehicle dynamics model and the track steering probability, so as to identify the dynamic risk range in the driving process according to the relative equivalent force. Themethod comprehensively considers the coupling relationship among human, vehicle and road elements, can be suitable for accurate identification of the driving risk range, and provides support for driving decision.

Description

technical field [0001] The invention relates to the technical field of safe driving of intelligent vehicles, in particular to a trajectory prediction-based driving risk identification method. Background technique [0002] In recent years, with the gradual increase of car ownership, the proportion of traffic accidents caused by cars has risen sharply, which has become a social issue that people focus on. Driving risk identification is the key to ensure driving safety during vehicle driving. Driving safety is related to the development of driver assistance systems (ADAS) and automatic driving, and also provides protection for the personal safety of drivers and passengers. Therefore, it is necessary to identify driving risks and accurately assess the risk status of surrounding vehicles and own vehicles. Risk identification is to define the range of influence in the driving process, and risk assessment can be understood as the possible damage to the vehicle during driving. The...

AI Technical Summary

Problems solved by technology

Although the effect of this analysis method is ideal, it focuses on specific scenarios, lacks comprehensive consideration of road environment and traffic elements, and is not suitable for wide application

Images