An Immunity inspired Real-time cooperative control framework for networked multi-agent systems

Abstract

This paper presents a cooperative control framework developed based on the inspiration from the immune system for controlling networked multi-agent systems. The framework is inspired from the meta-dynamics of lymphocyte repertoires in the adaptive immune system, including the continual circulation, continual turnover, concentration control and other relevant mechanisms. We design this framework for the control of a team of autonomous defending agents in RoboFlag Drill, a test-bed for studying cooperative systems. Simulation results are presented to demonstrate the effectiveness of the proposed immunity inspired cooperative control framework. The results of the simulations demonstrated that a set of defenders- can intercept attacker sets with larger set sizes from entering a specific Defense Zone for 60% of the randomly generated RoboFlag Drill problem instances.