Distributed multi-unmanned aerial vehicle relay network coverage method

The invention provides a distributed multi-unmanned aerial vehicle relay network coverage method, which comprises the following steps of: firstly, establishing a multi-unmanned aerial vehicle communication channel model, an energy consumption model and a collaborative trajectory optimization mathematical model, and converting into a Markov game model; secondly, a multi-agent deep reinforcement learning algorithm for centralized training and distributed execution is built, each unmanned aerial vehicle is provided with an action device and an evaluator, the action device outputs a flight path according to the state information of the unmanned aerial vehicle, and the evaluator outputs a corresponding action value to train the action device; and finally, constructing an action device with a dual-data-stream structure so as to improve the dynamic nature of the unmanned aerial vehicle. A layered multi-head attention encoder is constructed, an attention mechanism is utilized, the unmanned aerial vehicle can restrain