Switching control method of modular variable unmanned aerial vehicle system

The invention discloses a switching control method of a modular variable unmanned aerial vehicle system. Gain parameters are adjusted in real time by using approximation characteristics of an RBF neural network, and meanwhile, multiple self-learning methods are adopted in an RBF neural network learning algorithm to prevent a learning process from falling into local optimization. In this way, buffeting of the system can be effectively eliminated, and rapidity and robustness of the system are guaranteed. According to the method, a complex nonlinear system is converted into switching of a plurality of simple subsystems, different models correspond to the subsystems of the switching system, and model mutation caused by structural change is described by the switching law, so that the problem offailure of an original control law due to appearance mutation after aerial docking of the unmanned aerial vehicle is realized is solved. The unmanned aerial vehicle control law can be switched in time after butt joint of the