Quad-rotor unmanned aerial vehicle six-degree-of-freedom rapid collaborative fine formation method

The invention discloses a quad-rotor unmanned aerial vehicle six-degree-of-freedom rapid collaborative fine formation method, and the method comprises the steps: building a quad-rotor unmanned aerial vehicle aerodynamics and control model and a quad-rotor unmanned aerial vehicle collision box model, and training an unmanned aerial vehicle to carry out the collaborative flight path planning artificial intelligence based on a Kronecker factor trust threshold executor/reviewer deep reinforcement learning algorithm; and then designing a collaborative formation logic to convert the collaborative formation problem of the unmanned aerial vehicle group into a collaborative flight path planning problem of multiple unmanned aerial vehicle pairs, and finally realizing efficient and fine formation flight path planning of the four-rotor unmanned aerial vehicle group. The method is short in algorithm operation time, and safe, efficient, rapid and accurate unmanned aerial vehicle group cooperative formation flight path plan