DRGBT (Dynamic Rapidly-exploring Generalized Bur Tree) algorithm is intended for motion planning in dynamic environments. The main idea behind DRGBT lies in a so-called adaptive horizon, consisting of a set of prospective target nodes that belong to a predefined C-space path, which originates from the current node. Each node is assigned a weight that depends on relative distances and captured changes in the environment. The algorithm continuously uses a suitable horizon assessment to decide when to trigger the replanning procedure.
0 commit comments