This invention relates to the field of low-altitude airspace management technology, specifically to a method,
system, and storage medium for generating
noise maps and optimizing flight paths for low-altitude aircraft. The method includes: extracting deep environmental features through an environmental feature encoding network; calculating a real-time
noise map centered on the aircraft's current position using a physical-data
hybrid noise prediction network, combining a physical
acoustic model with neural network correction terms; generating a sequence of optimized flight actions for the future
time domain based on the
current noise map, aircraft state, and target point using a silent path optimization network; executing the first action in the action sequence, updating the aircraft state, and returning to perform rolling optimization for the next moment until the target point is reached. This invention, through the
deep integration of a
physical model and a data-driven model, ensures the
correctness of acoustic principles while utilizing neural networks to compensate for complex environmental factors, achieving high-precision, real-time noise distribution prediction.