Abstract: Multi-UAV systems transporting cable-suspended payloads in complex environments face multi-source risks from disturbances and obstacles. This paper proposes a risk-triggered distributed safety cooperative control method. The payload-induced disturbance is modeled and estimated via a disturbance observer. Collision risk, accounting for uncertainties in tracking, state estimation, and obstacle perception, is quantified using conditional value-at-risk. The proposed controller integrates robust formation control, collision avoidance, and disturbance compensation. Simulations and flight tests demonstrate improved control accuracy under payload disturbances and enhanced safety in complex environments, with reduced obstacle-avoidance conservatism.

Key words: Multi-UAV systems, risk quantification, safety collision avoidance control, anti-disturbance control, distributed cooperative control