Cooperative search is a fundamental task for the unmanned aerial vehicles (UAVs), and cognitive maps are necessary resources for the entire search process. However, due to the limited sensing range of an individual UAV, it naturally throws a tradeoff between the coverages and connectivity maintenance for the specific search implementations. A simple but efficient way to address this problem is to apply the formations. However, traditional methods using fixed formation structures and constant resolution maps are insufficient to improve the global search performances, due to their lack of adaptabilities and flexibilities. This paper proposes a novel method for the UAV formation searches that is based on the variable resolution cognitive map (VRCM) and the multi-branch tree (MBT) communication topology. The VRCM aims to establish the cognitive maps with different map resolutions to support the various search behaviors. An incremental Gaussian process (IGP) model is applied for the VRCM constructions, such that the behavioral diversity and search flexibility can be improved during the online manipulations. The MBT topologies are used to interconnect the UAVs where the root node is the formation leader that is capable of the plannings of the global search paths. A reconfiguration scheme based on the cognitive maps and benchmark formations is provided to optimize the coverage performance of the UAV formations. Simulation results demonstrate the advantages of the proposed method on the stability of switching formations, search accuracy, coverage efficiency, and retrospect behaviors to the areas with high target occurrences.

Developing a high-quality scenario library is crucial for evaluating more reliable autonomous driving systems. A fundamental prerequisite for constructing such a library is the generation of safety-critical scenarios. In this paper, we propose a nested game algorithm to assign trajectories and their time series to multiple background vehicles. This method aimed to generate safety-critical scenarios with varying degrees of interference for the tested vehicle. To ensure the realism of the generated scenarios, we extracted a series of natural trajectories from an existing dataset as the input for the algorithm. We then analyzed multiple types of scenarios generated by this method and evaluated their danger using generalized metrics, such as the Time-to-Collision (TTC) and Minimum Safe Distance Factor (MSDF), which demonstrated the effectiveness of our approach. The experimental results demonstrate that the nested game-based approach could efficiently construct safety-critical scenarios, contributing to the development of high-quality test scenario libraries.

With characteristics such as low cost, flexibility, and scalability, unmanned aerial vehicle (UAV) swarms demonstrate superior performance over single unmanned platforms and manned aircraft in search and surveillance missions. However, the conflicts between individual decisions and the tradeoff between search and connectivity in limited sensing areas still render UAV swarm search and surveillance inherently challenging. This paper proposes a subgroup differentiation-based decision-making framework for the UAV swarm, where two kinds of roles (informers and relays) are considered and each UAV can autonomously switch its role according to the task demand and environmental changes. The relay nodes provide larger communication scopes for connectivity preservation and contribute to the relaxation of informers’ constrained actions. In this way, the informing UAVs can maintain well-preserved explorations during the search process. The subgroup differentiation is based on a distributed framework where two sequentially linked auction-based operators are respectively developed for the action policies of informers and relays. The impact time control guidance is used for simultaneous arrival to realize the synchronous information fusion of swarm search findings. Simulation results demonstrate the efficient explorations, less conservativeness, and higher coverage efficiencies of the proposed method over existing advanced approaches in confined sensing environments.