A new Rapidly-exploring Random Tree (RRT) algorithm variant isproposed for dealing with time-dependent motion planning problemsin complex environments. The algorithm takes into account the time-varying nature of the goal region, the constraints, and the system dy-namics, and it provides solutions that are guaranteed to converge tominimal-time solution. The proposed algorithm is based on the originalRRT algorithm, and it utilizes a novel set of methods for generatingasymptomatically-optimal trajectories for systems subject to complexdierential constraints. The algorithm is evaluated in a simulation studyinvolving a Dubins car-like vehicle, which represents a crude descriptionof the dynamics of an aerial vehicle.
展开▼