Transforming Multi-agent Planning Into Single-Agent Planning Using Best-Cost Strategy

Abstract

This article presents a method to transform a multi-agent planning (MAP) problem into single-agent tasks through a goal delegation process using a best-cost strategy. This research work combines different properties of planning methods that describe the state of the art of MAP. The planning approach takes into consideration the cost associated to the action and the action sequence related to the set of goals that each agent must achieve. The proposed method was validated using the Graphplan planner taking into consideration the optimal solution in relation to cost, time and use of memory during the planning process. The results show the potential of the approach that can face the exponential growth-rate of planning time through the delegation of goals using a best-cost strategy to transform a MAP problem into as many as necessary single-agent tasks.