An exact solution approach for the mobile multi‐agent sensing problem
Gwang Kim, Jong‐Min Lee, Ilkyeong Moon · International Journal of Robust and Nonlinear Control · 2022
Abstract Multi‐agent systems are generally applicable in various fields and aim to coordinate the decisions from agents, each of which makes a local decision. In particular, research on operations management with mobile multi‐agents such as drones has gained prominence in recent years. In this article, we present a mobile multi‐agent sensing problem, which is formulated as a submodular maximization problem under a partition matroid constraint. When events detected by agents lead to severe and catastrophic consequences, obtaining (near)‐optimal solutions by using exact algorithms is crucial to reducing the probability of harmful situations. Therefore, in this article we propose an exact solution approach, using two valid inequalities for our problem to find a (near)‐optimal solution. Moreover, we deal with the risk‐averse decision and its cutting‐plane algorithm for our problem, which is to maximize conditional value‐at‐risk (CVaR). Finally, we show the performance of our algorithms through numerical experiments, including a case study on forest fires.