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.