Probabilistic path planning for multiple robots with subdimensional expansion

Glenn Wagner,Minsu Kang,Howie Choset

DOI: https://doi.org/10.1109/icra.2012.6225297

2012-05-01

Abstract:Probabilistic planners such as Rapidly-Exploring Random Trees (RRTs) and Probabilistic Roadmaps (PRMs) are powerful path planning algorithms for high dimensional systems, but even these potent techniques suffer from the curse of dimensionality, as can be seen in multirobot systems. In this paper, we apply a technique called sub dimensional expansion in order to enhance the performance of probabilistic planners for multirobot path planning. We accomplish this by exploiting the structure inherent to such problems. Sub dimensional expansion initially plans in each individual robot's configuration space separately. It then couples those spaces when robots come into close proximity with one another. In this way, we constrain a probabilistic planner to search a low dimensional space, while dynamically generating a higher dimensional space where necessary. We show in simulation that sub dimensional expansion enhanced PRMs can solve problems involving 32 robots and 128 total degrees of freedom in less than 10 minutes. We also demonstrate that enhancing RRTs and PRMs with sub dimensional expansion can decrease the time required to find a solution by more than an order of magnitude.