Çoklu iha'larla arama ve kurtarma: Hedef tespiti ve bağlantılılık

In this thesis, we present a multiple UAV system for SAR missions, focusing on efficient target detection, connectivity, and task assignment. The proposed framework addresses critical challenges in large-scale disaster scenarios, where UAVs must collaboratively detect, monitor, and maintain connectivity with multiple targets. To achieve these objectives, we employ two kinds of UAVs: search UAVs equipped with onboard sensors for rapid area scanning and relay UAVs that establish and sustain network connectivity among UAVs, targets, and GCS. Multiple target detection and connectivity strategies with varying levels of adaptability are proposed: (i) maintaining pre-planned paths while communicating detected targets to the GCS, (ii) deploying additional UAVs for monitoring detected targets without disrupting ongoing search operations, and (iii) dynamically re-planning UAV paths by assigning search UAVs to detected targets, providing the highest level of responsiveness to dynamic conditions. Our results demonstrate that less adaptive strategies achieve faster target detection but often compromise connectivity. Conversely, more adaptive methods significantly improve network performance at the expense of increased mission time. Integrating relay UAVs enhances connectivity by up to 60% compared to baseline planners, with the cost of increasing total mission time by up to 40% in specific configurations. An innovative connectivity-based target prioritization method is integrated into the proposed dynamic path planning algorithms for multi-UAV systems, customizing it for search and rescue missions. Our findings highlight the trade-offs between detection efficiency, network reliability, and overall mission time, providing valuable insights for deploying UAV systems in SAR missions.