Gain-scheduled and cascaded proportional–integral–derivative (PID) autopilots remain common practical baselines for fixed-wing unmanned aerial vehicles (UAVs), but training one shared learned controller for heading, altitude, and true airspeed across several maneuvers remains difficult. We study this problem under a strict reach-then-hold benchmark in which all the active channels must enter prescribed green bands and remain there for a terminal hold window. The proposed training recipe combines proximal policy optimization (PPO) with a tri-band maneuver-tracking reward and an outer bounded reward reweighting (BDR) step that updates the base reward weights from recent violation summaries under a Kullback–Leibler (KL) gate. In the JSBSim F-16 six-degree-of-freedom dynamics model, used here as a challenging surrogate benchmark for fixed-wing UAV autopilot learning, the learned controller transfers across a fixed five-lesson sequence, reaches strict success rates of 0.966 on turn and 0.921 on climb, and issues substantially smaller executed-command updates than the shared fixed-gain PID reference used here. Under the reported lesson sequence and step budget, fixed-weight PPO and a reweighting-only variant stall under the same envelopes, while speed remains the main bottleneck for both controllers. We further report exploratory long-horizon tracking, difficult-command stress checks, and an added command-filtered nonlinear dynamic-surface-control (CF-DSC) reference without retraining the learned policy. The CF-DSC results confirm that advanced non-reinforcement-learning (non-RL) controllers can be strong reference methods; therefore, within this reported simulator setup, BDR should be read as a practical and inspectable reward-scheduling heuristic for shared triad tracking rather than as a proof of superiority over all classical, nonlinear, or model-based controllers.

As UAV popularity in the Internet of Things (IoT) soars, so does the obstacle avoidance associated with it. The classic co-directional avoidance strategy, where the avoidance direction aligns with the obstacle’s movement, faces challenges in accurately interpreting the obstacle’s motion intent and is susceptible to errors in obstacle perception and prediction. To achieve more active and safe avoidance of dynamic obstacles by UAVs, this paper proposes a dual-layer obstacle avoidance path replanning framework. The upper-layer dynamic obstacle bypass decision model based on Transformer first applies a geometric constructor to explicitly generate key geometric features. It then interprets the movement intent of dynamic obstacles and performs spatial reasoning to actively choose a safer direction for bypassing obstacles. The lower-layer employs a direction-inspired path planning method. It first rapidly explores an initial safe path based on the decision direction, and then applies a combinatorial optimization techniques to achieve path smoothing and curvature constraint satisfaction. Finally, simulation experiments show that the proposed model can interpret the movement intent of obstacles and perform spatial reasoning to decide safe bypass direction, achieving a prediction accuracy of 98.93% — a 24.54% improvement over the baseline Transformer model. Furthermore, obstacle avoidance replanning experiments display that proposed framework achieves comparable planning times to classical RRT, approximately 10^–1s, while effectively ensuring paths comply with the counter-directional avoidance strategy.