Large-scale solar-powered unmanned aerial vehicles possess the capacity to perform long-term missions at different altitudes from near-ground to near-space, and the huge spatial span brings strict disciplines for its attitude control such as aerodynamic nonlinearity and environmental disturbances. The design efficiency and control performance are limited by the gain scheduling of linear methods in a way, which are widely used on such aircraft at present. So far, deep reinforcement learning has been demonstrated to be a promising approach for training attitude controllers for small unmanned aircraft. In this work, a low-level attitude control method based on deep reinforcement learning is proposed for solar-powered unmanned aerial vehicles, which is able to interact with high-fidelity nonlinear systems to discover optimal control laws and can receive and track the target attitude input with an arbitrary high-level control module. Considering the risks of field flight experiments, a hardware-in-loop simulation platform is established that connects the on-board avionics stack with the neural network controller trained in a digital environment. Through flight missions under different altitudes and parameter perturbation, the results show that the controller without re-training has comparable performance with the traditional PID controller, even despite physical delays and mechanical backlash.

Unmanned aerial vehicle (UAV) swarms offer unique advantages for area search and environmental monitoring applications. For practical deployments, determining the optimal number of UAVs required for a given task and defining key performance metrics for the platforms and payloads are crucial challenges. This study aims to address mission planning and performance optimization for cooperative UAV swarm search scenarios. A new clustering algorithm is proposed, integrating enhanced clustering techniques with ant colony optimization, particle swarm optimization, and crow search optimization. This jointly optimizes and validates the UAV numbers and coordinated trajectories. Sensitivity analysis and indicator optimization further examine specific scenarios to quantify platform and sensor factors influencing search efficiency. Lastly, sensitivity analysis and performance indicator optimization are conducted in specific scenarios. The modular algorithmic components and modeling techniques established in this work lay a foundation for continued research into real−world mission−based swarm optimization.

Multi-UAV cooperative encirclement tasks have attracted considerable attention in areas such as military defense and target interception. Fixed-wing UAVs face substantial challenges due to intrinsic dynamic limits, including their minimum velocity and turning radius, particularly when engaging evasive target and navigating in obstacle environments. This paper presents a hybrid deep reinforcement learning approach, in which a cooperative task environment is developed for fixed-wing UAVs that jointly integrates encirclement and obstacle avoidance. A composite MADDPG framework enhanced with curriculum learning is designed, employing progressive task staging and reward optimization to accelerate convergence and improve policy stability. Simulation results demonstrate that the proposed method achieves single-step encirclement success rates exceeding 80% in complex environments, while maintaining 10-step success rates around 70%, thereby strengthening both encirclement capability and obstacle avoidance safety in fixed-wing UAV swarm. This study provides new insights into the intelligent cooperative control of fixed-wing UAVs in high-risk missions.

Intelligent swarm is a powerful tool for targeting high-value objectives. Within the Anti-Access/Area Denial (A2/AD) context, an unmanned aerial vehicle (UAV) swarm must leverage its autonomous decision-making capability to execute tasks with independence. This paper focuses on the Suppression of Enemy Air Defenses (SEAD) mission for intelligent stealth UAV swarms. The current research field mainly faces challenges in fully simulating the complexity of real-world scenarios and in insufficient autonomous task planning capabilities. To address these issues, this paper develops a representative problem model, establishes a six-tier standardized simulation environment, and selects the Deep Deterministic Policy Gradient (DDPG) algorithm as the core intelligent algorithm to enhance the autonomous task planning capabilities of UAV swarms. At the algorithm level, this paper designs reward functions corresponding to UAV swarm behaviors, aiming to motivate UAV swarms to adopt more effective action strategies, thereby achieving autonomous task planning. Simulation results demonstrate that the scenario and architectural design are feasible and that artificial intelligence algorithms can enable the UAV swarm to show a higher level of intelligence.

Maritime security is vital to national security. The application of unmanned aerial vehicle (UAV) clusters in marine anti-submarine warfare (ASW) presents a new and significant challenge worthy of in-depth study. Based on the anti-submarine principle of geomagnetic anomaly detection and the Find-Fix-Track-Target-Engage-Assess (F2T2EA) framework, this paper divides UAV cluster ASW operations into two stages: regional coverage and cooperative convergence. The regional coverage stage enables the UAV cluster to perform a broad search for submarines, while the cooperative convergence stage facilitates the precise positioning of detected submarines. In simulation, the combat scenario of five UAV clusters against a submarine is carried out. In the given area, the UAV can locate and stalk the target submarine in limited time. Simulation results demonstrate the feasibility of the proposed approach, providing reference for advancing marine ASW capabilities and related research.

Rapid wildfire response requires UAV swarms to approach a fire region from dispersed initial states, coordinate their arrival with limited information exchange, and adapt terminal deployment to updated fireline geometry. This paper proposes a two-stage distributed guidance framework for time-coordinated fireline deployment. In Stage I, a regularized closing-time variable is combined with bounded-degree topological interaction, proximity-based collision avoidance, and persistent longitudinal speed regulation to reduce rendezvous-time dispersion. In Stage II, the updated fireline is discretized into service points and assigned using a geometry-aware structured strategy, followed by distance-aware terminal synchronization and local safety control. Conditional analysis establishes convergence of the reduced Stage-I coordination dynamics under admissible switching topologies and sufficient separation conditions for ideal Stage-II deployment. In the 12-UAV Stage-I comparison, the proposed method reduces the arrival-time spread to 1.34 s, compared with 16.12 s for fixed-topology cooperative guidance and 4.20 s without longitudinal timing correction. Stage-I tests are further extended to 100 UAVs, while Stage-II assignment is evaluated with up to 60 UAVs and maintains low computational cost. In matched three-dimensional tests involving measurement noise, wind disturbance, and bounded communication delay, the proposed method achieves the smallest terminal timing error among the compared strategies, with a mean Stage-II spread of 2.654 s over the four tested conditions. The results demonstrate effective and scalable time-coordinated deployment under the considered kinematic and communication conditions.