Local path planning based on improved Dynamic window approach

Abstract

An improved dynamic window method is designed as a local path planning method. Aiming at the problem of unreasonable path selection in traditional dynamic window approach when it is close to the target point and dense obstacles, an improved dynamic window approach is proposed. A new objective function is introduced in the algorithm to improve the speed selection when the robot approaches the target point. The number of obstacles is introduced into the velocity evaluation function to improve the problem that dynamic window approach is easy to bypass dense obstacles. Simulation experiments in different environments show the effectiveness of the improved algorithm.