An extended area-partition-involved collaborative patrolling game in chemical clusters considering attackers’ bounded rationality and parameter uncertainty

Abstract In our previous study—area-partition-involved collaborative patrolling (APCP) game, we proposed a cooperative patrolling game by utilizing multiple patrol vehicles to protect the chemical cluster. Except for fixed countermeasures like intrusion detection, the APCP game was proved to reduce the security risks of chemical clusters being attacked by terrorists effectively. However, some general assumptions (i.e., complete information game and perfect rational players) in the game model make it difficult to be applied in reality since the game model is not robust enough to deal with some common uncertainties on attackers. In this paper, we extend the APCP game model by considering the uncertainties of the game model from two aspects: (1) One is the bounded rationality of attackers. Attackers may adopt the e-optimal response strategy without maximizing their payoffs strictly. In this circumstance, we propose the bounded rational APCP game solver. (2) The other is the uncertainties of the defender’s estimation on the attacker’s parameters. Some parameters regarding attackers are changed from exact values to distribution-free intervals. Correspondingly, we propose the interval APCP game solver. Through extensive simulations on a chemical cluster (i.e., the Antwerp Chemical Cluster), results of the case study indicate that when considering two types of uncertainties mentioned above, the defender’s payoff will be slightly reduced, whereas the more robust patrolling routes are more easily convinced and adopted by the management department of chemical clusters and improve the process safety of chemical clusters.