Optimal Distributed Market-Based Planning for Multi-Agent Systems with Shared Resources
Sue Ann Hong, Geoffrey Gordon; JMLR W&CP 15:351-360, 2011.
AbstractMarket-based algorithms have become popular in collaborative multi-agent planning due to their simplicity, distributedness, low communication requirements, and proven success in domains such as task allocation and robotic exploration. Most existing market-based algorithms, however, suffer from two main drawbacks: resource prices must be carefully handcrafted for each problem domain, and there is no guarantee on final solution quality. We present an optimal market-based algorithm, derived from a mixed integer program formulation of planning problems. Our method is based on two well-known techniques for optimization: Dantzig-Wolfe decomposition and Gomory cuts. The former prices resources optimally for a relaxed version of the problem, while the latter introduces new derivative resources to correct pricing imbalances that arise from the relaxation. Our algorithm is applicable to a wide variety of multi-agent planning domains. We provide optimality guarantees and demonstrate the effectiveness of our algorithm in both centralized and distributed settings on synthetic planning problems.