Maximal operation time estimation for modular and self-reconfigurable robots with output current constraints

A key problem in modular and self-reconfigurable robot power sharing research is that of estimating the globally maximal operation time (MOT) of a system of connected modules that share power with one another, as this information is vital to determining resource allocation in power sharing schemes. Existing MOT estimation algorithms do not consider the effect that output current constraints at each module — i.e., the fact that each module can only share a certain amount of power with other modules — have on the estimation of MOT information. This paper proposes both centralized and distributed algorithms for estimating the MOT of systems of connected modular and/or self-reconfigurable robots in which each robot module is subject to constraints on the amount of electrical current it can output. These algorithms are based on a transformation of the power sharing problem (with output current limits) to a minimum-cost flow problem, for which efficient algorithms exist. The proposed algorithms are validated in large-scale simulations to demonstrate their correctness, feasibility, and scalability.