Human-Aware Path Planning with Improved Virtual Doppler Method in Highly Dynamic Environments
Kuanqi Cai,Weinan Chen,Chaoqun Wang,Shuang Song,Max Q. -H. Meng
DOI: https://doi.org/10.1109/tase.2022.3175039
IF: 6.636
2023-01-01
IEEE Transactions on Automation Science and Engineering
Abstract:Human-aware path planner is essential for achieving harmonious coexistence between humans and robots in highly dynamic environments. In this paper, we propose an integrated framework to find the optimal path in the complex environment with considering collision risk, social norms, and crowded areas. In the proposed framework, a general dynamic group model (g-space) based on the Gaussian Mixed Model (GMM) is proposed as the social norms of dynamic groups, which not only considers the factors of humans (e.g., pose, quantity, distribution, psychology) but also establishes the proximity and human interacting constraints of dynamic groups. An integrated Collision Risk and Human Space (CR&HS) model is applied to achieve human-acceptable behaviors, in which both collision avoidance, human comfort, and interference-free constraints have been involved. Moreover, an Improved Virtual Doppler Method (IVDM) has been used to realize safety navigation to avoid the robot falling into the crowded area. Finally, the proposed framework has been utilized with the sampling-based rapidly-exploring random tree. Experimental results demonstrate that the proposed method can generate the optimal human-aware collision-free path in complex environments. Note to Practitioners—This paper aims to plan an optimal trajectory for the robot in highly dynamic environments. In this field, it is still a challenging task to plan a trajectory with collision-free, human-aware, and crowd-aware. To do that, we present an integrated framework to generate the optimal trajectory by involving the collision risk, social norms, and human density. First, the g-space model is adopted as interference-free constraints of dynamic groups. The integrated knowledge fusion model (CR&HS) then penalizes the manners which have higher collision risk and adverse effects on human interaction or human comfortable. Besides, human motion and density are provided to a robot by IVDM. The proposed framework is utilized in the sampling-based rapidly-exploring random tree as the evaluation module. Finally, the feasibility and reliability of the proposed method have been verified by experiments in different simulated environments. The proposed framework can be applied in most mobile service robots to achieve human-friendly manners.