Research On Quadrilateral Aircraft Control Algorithm Based On Multi-information Fusion

Quadrotor is a vertical take-off and landing of unmanned aerial vehicles machine,which can remote control,or autonomous flight of aircraft.Following the constant development and ripeness of sensor technology,embedded technology and control algorithm,Quadrotors utilized widely in military,civil,scientific and commercial fields due to the characteristic of low cost,high performance,small size,strong survival ability and diversity flight mode.Quadrotor system,in which involved are electro mechanics,aerodynamics,navigation principle,principle of automatic control and communication technology,has fully reflected the features of diversity and comprehensive and drawled more enthusiasts' attention.Selecting Quadrotor as the research object,the paper describes the summary of research status and purpose of researched development status.It regards the control algorithm as the theme and with multi-sensor information fusion strategy to validate the effectiveness by using Simulink and actual experimentThis thesis mainly contributes to the following four aspects.Firstly,on the base of introducing mechanic structure and flight theory,this paper establishes geographic and body coordinate system according dynamic characteristic.According the Newton's second law and the law of momentum,the aircraft translational and rotational equation are given and on the basis of above all,it combines correlating parameter with flight characteristics of QBallX4 unmanned aerial vehicle test platform to simplify aircraft mode rationally.Secondly,according to mathematical mode,double close-loop controller has been taken based on individual characteristic of internal and external loop.For the purpose of facilitating implementation in engineering and integrating lots of indexes which include smooth and celerity of controller,inner loop is designed using LQR control algorithm.To overcome the traditional problem of poor performance in suppressing noise,feedback of velocity measurement is introduced in input terminal and outer loop is designed using differential forward PID control algorithm.Nonlinear constraint is given as there are transformational relations between control variable of inner and outer loop.Flight control system is built to verify the control algorithm by simulation.Then,for so many problems on the metrical data of multi-sensors in Quadrotor system,introducing related contents about information fusion.Information of position and attitude fusion architecture is proposed specific to the sensors of QBallX4 aircraft,with the improved Sage-Husa Adaptive Kalman Filter,in order to get the bettrer optimal position and attitude of the estimate.Fourth,a series of flight tests are done onQBallX4 experiment platform,to validate the aircraft alg,orithm and sensor information fusion strategy,In the experiment,verify the validity of the algorithm,aircraft flying fixed-point hovering and fixed point is implemented by adjusting related parameter.All of above have verified the feasibility and preferable performance of double closed-loop controllesr designed in this paperFinally,the paper summarizes the whole working condition and looks forward to future work.