The increasing cost tree search for optimal multi-agent pathfinding

Sampling-Based Motion Planning Algorithms for Replanning and Spatial Load Balancing

“…A dynamic programming scheme is used in [41] to develop an online coordinated motion planner. While the algorithms in [56] and [44] use graphs, they do not use samplingbased motion planners. Sampling-based motion planners such as the RRT are used in [62] and [52].…”

Monte-Carlo Fork Search for Cooperative Path-Finding

“…ICTS [14] is a two-level search: a global level and a low level. At the global level, the search proceeds on an Incremental Cost Tree (ICT) in which one node corresponds to a vector of costs.…”

“…In most work [16], [14], [15], [10], [7], the optimized target is SUM. However, in this paper, we optimize MAX, that is to say the global elapsed time.…”

“…actualSeq is the sequence of positions played at each iteration. root is the root node initialized with the starting position a. NMCFS is a while loop (lines [7][8][9][10][11][12][13][14]. it is the number of iterations to perform.…”

“…A* with operator decomposition (A*+OD) [16] is a speed-up version of A*. ICTS [14] searches a solution in an Incremental Cost Tree (ICT). The weakness of the coupled approach is its inability to solve large problems or including complex coordination between agents.…”