Hybrid Intelligent Algorithms Enhanced Optimization of Approach Trajectories for Low-Altitude Flight Targets in Dynamic Thunderstorm Scenarios

The instrument approach is the segment most severely affected by convective weather during the entire flight. This study introduces an integrated method for safe four-dimensional (4D) approach-trajectory optimization in dynamic thunderstorm scenarios. The process begins by extracting structural features of thunderstorm data through clustering and Monte Carlo-based correlation analysis, followed by short-term movement prediction using weighted interpolation. Based on the above results, a hybrid RRT-APF algorithm generates a dynamic thunderstorm-avoidance trajectory. To address local distortions in the plan view, the method applies a round mean filter to remove unstable waypoints while preserving key spatiotemporal characteristics of the route. Ultimately, a flight-procedure validation system assesses the overall research performance. The results demonstrate a thunderstorm-movement prediction accuracy of 93.7%, an optimized descent gradient of 3.73% during the final approach, and the hybrid RRT-APF algorithm cost of 0.2795, in compliance with the instrument approach design standards. Collectively, the research method yields smooth and reliable 4D trajectories, providing an effective basis for safe approach operations in convective weather, as well as future research on multi-aircraft conflict resolution under thunderstorm conditions.