Adaptive Vector Field Histogram Plus method for local path planning under human—robot collaboration

Abstract

As an expanding field that caters to the needs of the elderly and people with disabilities, intelligent wheelchairs aim to enhance the efficiency and safety of individuals with disabilities. Path planning, a pivotal technology under intelligent wheelchair systems, has emerged as a focal point of scholarly exploration. This paper proposes advancements in the vector field histogram plus (VFH+) algorithm, thereby introducing the adaptive VFH+ algorithm (AVFH+). Notably, this paper redefines the calculation of obstacle density and introduces the concept of custom target point density. By dynamically modulating the threshold, we effectively resolve the problems of oscillatory paths during obstacle avoidance, thus averting the predicament of the robot. In addition, the robot's speed is adaptively adjusted based on the opening width, and the trajectory is smoothed using Bezier path planning, enabling the robot to reach the target position through a shorter collision-free and smooth path. The improved AVFH+ algorithm demonstrates excellent path planning capability through multiple simulation experiments in various map environments with different starting points and target locations. The algorithm has the ability to adapt to complex environments, accurately perceive and avoid obstacles, and exhibits significant generalization capabilities. This makes it significant in dense area path planning, providing safer and more efficient mobility solutions in areas such as smart wheelchairs, and actively contributing to the quality of life of people with disabilities. Future research could explore the consideration of dynamic obstacles and validate the performance of the algorithm in real-world environments.