Dynamic Task Allocation Method for Heterogenous Multiagent System in Uncertain Scenarios of Agricultural Field Operation

Abstract

This study focuses on the problem of dynamic task allocation for a heterogeneous multiagent system (MAS) in uncertain scenarios and its application in agricultural field operation. Previous studies lacked robustness or efficiency for uncertain environments especially in agricultural field, such as agent removal, agent inclusion, changes of agent capabilities, and task changes. We present herein a novel concept of the potential field of capability influence (PFCI), and based on which, we can estimate potential overloaded tasks. This provides an opportunity to improve the allocation of the remaining tasks. Then, we propose a heuristic-based clustering auction (HBCA) method by introducing PFCI into the auction mechanism to achieve an effective and efficient task allocation. The whole method consists of the three following phases based on the auction mechanism: 1) auctioneer preprocessing phase that adopts a capability priority strategy to select the preferable agent; 2) bidder preprocessing phase that introduces the PFCI to obtain a list of clustered tasks heuristically; and 3) adaptive auction phase that applies the auction mechanism to achieve an appropriate match between agents and tasks. Numerical simulations with different uncertain field operation scenarios validate the effectiveness of the proposed method. The results show that the proposed HBCA can dramatically reduce the total assignment cost and outperform state-of-the-art methods, especially in the case of obvious variances in agent capabilities.