Research On Cloud Model Parameter Self-tuning PID Control Of Quadrotor UAV

With the rapid development of AI technology,intelligent control technology is widely used in the field of unmanned system control,quadrotor UAV is one of these systems.Unlike the UAV of the classic configuration,the quadrotor UAV can fly through simpler control methods due to its structural characteristics,but due to its four-input and six-degree-of-freedom characteristics,the system is statically unstable,with the characteristics of nonlinearity,under-actuation and strong coupling,quadrotor UAV is a typical uncertain nonlinear system.The cloud model control theory is based on the use of computational data to simulate the artificial control effect as the main idea.In this case the model of the accurately controlled object is not needed,the cloud model qualitative reasoning method is adopted to achieve the control requirements of the complex nonlinear system.It not only preserves the controlled objects and the uncertainties in the environment,but also has good adaptability and robustness to system state parameters and state changes,and has powerful nonlinear processing capabilities.The innovation of this paper is to apply the cloud model theory to the quadrotor UAV control system,and introduce the genetic algorithm to self-tuning the driving factors of the two-dimensional composite cloud model.The work of this paper is carried out in the following aspects:(1)Nonlinear dynamics modeling.The Euler equation and Newton’s laws of mechanics are used to analyze the flight principle of the quadrotor UAV.The aerodynamic and torque characteristics are derived by the three-line motion and the three angular motions of the six-degree-of-freedom motion of the quadrotor UAV.The nonlinear full-state feedback equation of the quadrotor UAV is obtained,which lays a foundation for further study of the control algorithm.(2)Design a quadrotor UAV control system and a one-dimensional composite cloud model controller.According to the nonlinear dynamic model of the quadrotor UAV,the internal and external loops and the method of increasing the virtual control amount are used to decouple the model,and the control system needed for the experiment is designed.The design of the inference mapping algorithm based on the one-dimensional cloud model is designed.A one-dimensional composite cloud model controller was used and related simulation experiments were carried out.(3)Design a quadrotor UAV two-dimensional composite cloud model controller.The actual working environment of the quadrotor UAV has the influence of uncertainty disturbance.For the problem that the one-dimensional composite cloud model controller has poor control effect in the presence of external disturbance,a two-dimensional cloud model inference mapping algorithm is introduced to design a kind of two-dimensional composite cloud model controller and simulation experiments in relevant hypothetical environments.(4)Design a two-dimensional composite cloud model controller optimized by genetic algorithm.Due to the strong coupling characteristics of the quadrotor UAV,it is difficult to achieve the optimal control effect by relying on the knowledge and debugging of the cloud model driving factor.In order to realize the true cloud model parameter self-tuning controller,the introduction of genetic algorithm theory to the design the two-dimensional composite cloud model controller is optimized,the genetic algorithm is used to calculate the driving factor,and relevant simulation experiments are carried out.