A Vector Field Histogram Based Fuzzy Planner for Robot Navigation in Dense Crowds

Real-time planning is an essential requirement for mobile robot navigation in dense crowds. This paper proposes a fuzzy planner using a vector field in polar coordinates for robot navigation in a human-robot coexisting environment. The proposed planner constructs a vector field histogram in polar coordinates based on the observed obstacles. An objective function is proposed to get an optimal sector among the vector field histogram. Finally, a fuzzy inference system is proposed to calculate the linear and angular velocities of the robot based on the value of the selected sector and the current robot heading. The performance of the proposed planner is evaluated in simulations and real-world experiments using a SUMMIT mobile robot with a Microsoft Kinect RGB-D camera and a HOKUYO-10LX laser sensor. In both simulations and real-world experiments, the results show that the proposed method enables the robot to reach its destination with safe and efficient motion. The computing time of the proposed planner in each planning period is only 1.4e
⁻⁴
seconds, which is suitable for mobile robots with a low computational cost.