Position And Attitude Control Of A Quadrotor Aircraft Based On Tensor Product Fuzzy Modeling

The quadrotor aircraft has simple mechanical structure and is widely used in im-portant fields such as plant protection,patrol inspection and exploration.It can not only perform tasks in a narrow environment,but also has the advantages of easy carrying and low later maintenance cost.It also involves the cross integration of many sophisticated disciplines,so it has a very wide range of research value.Taking the nonlinear model of quadrotor aircraft as the research object,a double closed-loop control strategy for the complexity of high-order system control of quadrotor aircraft with good robustness is proposed to realize the tracking control of position and attitude system.The main work of this thesis is summarized as follows:1)Firstly,the pseudo linear parametric model of non-equilibrium point lineariza-tion of quadrotor aircraft is analyzed and established in this thesis,and the mathematical model of the controlled object is given,which provides a research object for the subse-quent controller design and analysis.2)According to the strong coupling and nonlinear characteristics of the high-order system of quadrotor aircraft,quadrotor aircraft’s model is decoupled into position loop subsystem and attitude loop subsystem.The classical isometric sampling method can not accurately obtain all the information of the nonlinear entry function in the ten-sor product model of attitude loop subsystem,a local extrema refinement-based tensor product model transformation with parallel distributed compensation control method is proposed.The simulation results show the effectiveness of the control algorithm.3)A novel double closed-loop control strategy for quadrotor aircraft model is pro-posed,that is,the fully-actuated control method and the idea of pole placement to fur-ther is used in design the feedback control law in the position loop subsystem;The parallel distributed compensation control method based on tensor product model trans-formation is applied in the attitude loop subsystem.The classical isometric sampling method can not include all the information of the quadrotor aircraft model,a parallel distributed compensation control strategy based on varying-input method is proposed,on which the uniform design sampling method is added,the simulation results show the effectiveness of the parallel distributed compensation control based tensor prod-uct model transformation under the introduction of varying-input method and uniform design sampling method.4)A global optimization controller of quantum bacterial foraging algorithm is proposed for solving the problem of adaptive parameter adjustment under parallel dis-tributed compensation control based on tensor product model transformation.The al-gorithm can ensure that the control parameters reach the optimal solution in the feasible region,which lays a foundation for the subsequent virtual reality simulation platform and physical verification platform.5)A virtual reality simulation platform is built through MATLAB/SIMULINK to test the effectiveness of the proposed double closed-loop control strategy;A physical verification platform with parrot minidrone aircraft as the core is designed,and fully verify the effectiveness of the double closed-loop control strategy and the necessity of global optimization controller through indoor and outdoor experiments.