Multi-UAV cooperative search and coverage control in post-disaster assessment: experimental implementation

Abstract

This paper provides simulation studies and experiment validation of cooperative coverage control of a multi-UAV system to quickly and initially assess the environment immediately after an earthquake. Quick response of disaster management team in the early hours after the earthquake can play a significant role in reducing the degree of damages and casualties. We consider an earthquake disaster scenario wherein four fixed-wing UAVs initially fly over the affected area and perform a quick general scan to identify critical areas wherein people have suffered severe injuries or buildings have been destroyed. The multi-rotors then fly over the critical areas identified by fixed-wing UAVs, as close as possible, to extract more detailed information and exactly localize the victims and survivors or also damaged infrastructure as stationary targets in a cooperative search and coverage control problem. For the first phase of the mission, we simulate a cooperative multi-agent system comprising four fixed-wing UAVs. For the second phase of the mission, we experimentally implement cooperative search and coverage control of a multi-UAV system including three DJI Tello quadcopters to gather more detailed information about the critical section identified by fixed-wing UAVs in the first phase of the mission and find the exact location of the victims or survivors as well. In cooperative coverage control, agents communicate with each other and exchange information. Therefore, at every moment of the execution of the mission, all agents have a unique cognitive map of the environment. The comparison of cooperative and non-cooperative methods shows that the uncertainty of the environment reaches the minimum acceptable value much earlier and the maximum coverage of the area will be achieved much faster in the collaborative distributed control method.