Abstract:
A problem of path planning for an agent moving amidst static and dynamic obstacles is considered in the paper as a graph search problem. A novel method to solve it is introduced. The method is based on the combination of two complimentary ideas: any-angle path finding, i.e. augmenting the graph with the edges that connect initially non-adjacent vertices, and safe interval path planning, i.e. grouping the timesteps into the continuous intervals and operating them during heuristic search for a solution. Theoretical and empirical investigation of the algorithm are performed.