Online Robot Navigation Using Discrete Waypoints via Time-Varying Guidance Vector Fields

In this study, we introduce a new online robot navigation strategy that exclusively utilizes discrete waypoints, steering a robot with the capability to proficiently address two critical tasks: path-following, even along self-intersected paths, and pursuing a moving target at a predefined distance within a two-dimensional (2-D) plane. First, the thin-plate spline (TPS) interpolation algorithm is employed to transform waypoints into an analytical continuous target curve. Subsequently, a comprehensive navigation framework that simultaneously considers computational cost and the upper bound of interpolation errors is proposed. This framework has the capability to generate time-varying guidance vector fields (GVFs) as guidance signals, which facilitates the robot to smoothly converge to the target curve. Furthermore, the GVF can be extended to accomplish collision-free navigation in cluttered environments. Experimental results demonstrate the effectiveness of our proposed approach in real-flight scenarios.