Sampling Based Path Planning and Vector Fields for Curve Tracking by UAVs

This paper describes the combination of artificial, circulating vector fields based on implicit curves as proposed in Gonçalves et al. (2010) with a sampling based navigation technique in the same spirit of VFRRT (vector field rapidly exploring random tree) from Ko et al. (2014). The resulting two stage strategy outputs feasible paths considering a Dubins airplane model from Owen et al. (2015), that avoids known static obstacles in the three dimensional space. In simulations the proposed algorithm outperforms a pure sampling based approach.