W.J. Hu, B.L. Chen

Pages: 63-72

Abstract
In order to improve the detection accuracy and collision avoidance success rate of driverless vehicle moving obstacles, this paper proposes a new method for detection and collision avoidance of driverless vehicle moving obstacles. Firstly, the HDL-64E 3D laser radar is used to collect the obstacle data, calculate the similarity of the moving obstacle, obtain the coordinates of the moving obstacle, and construct the detection function of the moving obstacle; Secondly, the force balance model of the abacus is established by using the road abacus model; Finally, under the constraint of the force balance model, the optimal steering wheel angle of the driverless vehicle is calculated to complete the collision avoidance of the driverless vehicle. The experimental results show that this method can improve the detection accuracy and anti-collision accuracy of moving obstacles, and the average anti-collision accuracy reaches 99.5%.
Keywords: driverless vehicle; movement obstacle detection; collision avoidance method; road abacus model