A Novel Distributed Fault-Tolerant Cooperative Control Approach for Auto-Reconfiguration of a Multi Agent System in Natural Disasters

This paper presents a novel Fault Tolerant Cooperative Control (FTCC) scheme for automated victim search in a flooding disaster scenario using a cooperative multi agent system. One of the most common reasons of UAVs failure is unfavorable weather condition which can be so harmful in climate-related disasters. Small UAVs especially quadcopters are highly vulnerable to bad weather condition which can be led to failure of one or more of them in a multi-UAV system. Thus, cooperative coverage control of a multi agent system have to be designed robust to the probable failures. In our innovative algorithm, when one or more of agents are outgoing, the closest agent which its surrounding has less priority than the failed agent's neighboring cells, decides to do auto-replacement and take up the target position of the failed agent. After failure occurred, the remaining agents calculate the virtual path of the failed UAV and its virtual path will be the basis of remaining agents’ computations, as before. Then, if each of the remaining UAVs is the nearest one to the outgoing agent’s next virtual position, it has to decide to do auto-replacement if the target probability or uncertainty of virtual agent’s surrounding is more which makes it with higher priority. The results indicate that our novel FTCC algorithm will efficiently compensate the lack of lost agent.