Dynamic task allocation based on auction in robotic mobile fulfilment system

Abstract

Task allocation is one of the important factors affecting the efficiency of robotic mobile fulfilment system (RMFS). In this paper, a dynamically changing task allocation model is constructed, with the overall maximum profit as the optimization objective, and allows robots that are performing tasks to participate in the task allocation. Using the auction algorithm, three dynamic allocation strategies are developed: queued allocation strategy (QAS), immediate allocation strategy (IAS), and reservable allocation strategy (RAS). This paper conducts simulation experiments to compare and analyze the proposed three dynamic allocation strategies, static allocation strategy (SAS) as well as a heuristic algorithm (HAS). Simulation results show that RAS, with robots' tasks changing dynamically, is better at increasing the number of picking orders and reducing the distance travelled by robots than other proposed strategies, which improves the picking efficiency of RMFS.