Research On UAV Path Planning And Control In Complex Environment

Quadrotor UAVs have advantages of flexible structure,simple mechanical structure,and easy maintenance,and are widely used in agricultural,power,photographic and other fields.In order to further promote the application of quadrotor UAVs and enhance their operational capabilities,two core issues that need to be focused on are how to design an efficient path planning algorithm and improve the accuracy of trajectory tracking.At present,most of the research space for path planning algorithms is two-dimensional plane,while relatively few studies have been done on path planning in three-dimensional space.In addition,quadrotors are under-actuated,strongly coupled,and nonlinear MIMO systems,making it difficult to establish accurate mathematical models,and thus present certain difficulties in the design of controllers.Based on the above issues,this paper mainly studies the following:First,based on the construction of quadrotor and the basic principles of flight fundamentals,the transformation matrix of coordinate system is deduced and the rigid dynamics model of quadrotor is established with the relevant theories as support;The principles of digital map and grid map construction are discussed,and after comparing and analyzing their advantages and disadvantages as the workspace model of quadrotor,the grid map is selected as the environmental map model for subsequent research.Secondly,by analyzing and summarizing several classic graph search algorithms,considering the problem of global static algorithms in UAV path planning unable to dynamically avoid obstacles and the path constraints due to UAV dynamics,Theta~* algorithm is applied for global path planning,and dynamic window approach is improved to track path nodes for quadrotor path planning in a three-dimensional environment.By building the path planning algorithm for quadrotor in MATLAB,a safe and smooth trajectory curve satisfying the dynamics constraints can be obtained.Finally,the control principles of the sliding mode controller and the model predictive controller are expounded.According to the characteristics of the quadrotor system,the model predictive controller,the ordinary sliding mode controller,and the fast non-singular sliding mode controller are designed.The simulation model of the quadrotor is built on the SIMULINK,and the three controllers are respectively used for trajectory tracking simulation verification.All three controllers successfully achieved tracking,among which the fast non-singular sliding mode controller has the advantages of less computation,strong anti-interference ability,and fast response compared with the other two controllers.Through the construction of the quadrotor UAV flight test platform,the real reliability of the proposed control algorithm is further verified.