Research On Control Algorithm Of Quadrotor Aircraft Based On MPC

In recent years,in the field of robotics,quadrotors have attracted attention at home and abroad due to their advantages of convenient operation,simple structure,and good maneuverability,and have been widely used in various fields.Since the quadrotor is a highly coupled and underactuated control object,the research on its control algorithm faces many challenges.For this reason,this paper takes the small quadrotor aircraft as the research object,focusing on dynamic modeling analysis,quadrotor pose controller state space model and step response model,MPC controller path planning and simulation,physical flight test,etc.The main research work is as follows.This paper firstly analyzes the force of the quadrotor aircraft structure,and establishes the dynamic model of the airframe through the Newton-Euler principle,measures and calculates the relevant parameters through the system parameter identification,and passes the strongly coupled nonlinear dynamic equation through the series expansion and other methods are linearized,and establishes two expressions of state space and transfer function.Then,according to the dynamic equation of the quadrotor,the model predictive control algorithm based on the state space model and the step response model was designed,and the simulation experiment of the position and attitude of the quadrotor was carried out by using Matlab＆Simulink software.Based on the advantages,disadvantages and usability of these flight control algorithms,by comparing the two control methods,the controller based on state space with better control effect is selected as the basis for the design of the quadrotor flight controller,so as to achieve better flight control effect.Then,in order to verify the tracking effect of the quadrotor aircraft control algorithm,starting from the path planning principle,the path tracking of RRT,RRT* algorithm and MPC controller is studied.The experimental comparison and simulation of traditional RRT and RRT* in fixed obstacles and random obstacle environments are carried out.The two-dimensional and three-dimensional paths planned by the RRT* path planning algorithm are processed by smoothing and softening as the reference input of the quadrotor aircraft.The two-dimensional and three-dimensional space path tracking simulation tests are carried out,and the analysis of the tracking error shows that the designed predictive controller based on the state space model has faster tracking performance.Finally,starting from the hardware construction of the quadrotor aircraft and the software design of the control system,the physical platform of the aircraft is designed.Among them,the open source flight control PX4 is used as the basic platform for secondary development,the upboard is used as the controller algorithm processing platform,and a set of ground monitoring software is independently developed for various flight data and status display and monitoring.The designed control system is embedded in the Prometheus open source platform,and simulation flight and physical flight experiments are carried out to verify the feasibility and engineering feasibility of using the MPC algorithm as the control algorithm of the quadrotor aircraft.The research in this paper provides a certain reference value for the control verification and engineering realization of the model predictive control algorithm on the quadrotor aircraft.