Goal state optimization algorithm considering computational resource constraints and uncertainty in task execution time

A search methodology with goal state optimization considering computational resource constraints is proposed. The combination of “an extended graph search methodology” and “parallelization of task execution and online planning” makes it possible to solve the problem. The uncertainty of the task execution time is also considered. The problem can be solved by utilizing a random-based and/or a greedy-based graph-searching methodology. The proposed method is evaluated using a rearrangement problem of 20 movable objects with uncertainty in the task execution time, and the effectiveness is shown with simulation results.