Real-time Collision Avoidance Motion Planning using Vehicle Lateral Angular Velocity Constraints
Sung-Woo Heo, Tae Hyoung Park, VALUE VALUE, VALUE VALUE · Journal of Institute of Control Robotics and Systems · 2019
This paper proposes a method for motion planning of a vehicle considering lateral angular velocity constraints. Motion planning is very important for the safe travel of an autonomous vehicle. In this work, the steering angle of the vehicle is included in the cost function in order to achieve smooth operation of the vehicle. and path optimization uses dynamic programming to select the paths that satisfy the constraint with the least cost. After selecting the optimal path, the longitudinal velocity profile is determined using a fuzzy speed controller. The steering input of the vehicle and the distance from the obstacle to follow the path are used as the input of the fuzzy speed controller, and the controller provides longitudinal acceleration and deceleration of the vehicle as the output. Experiments were conducted for three scenarios using vehicles equipped with Lidar, IMU, GPS, and cameras. Experimental results show that the proposed motion plan generates a path within the steering constraint condition of the vehicle and runs effectively.