Research On Attitude Estimation And Control Algorithm For Rotorcrafts

As one kind of important vehicles used in military and civilian applications,Unmanned Aerial Vehicles(UAVs)have gained widespread focuses these years.With the development of science and technology,UAVs are experiencing the tendency to become smaller,smarter,and more versatile.Due to their advantages such as the simplicity of mechanical structure,high reliability and maneuverability,capability of vertical take-off and landing(VTOL),rotorcrafts obtain the dominant applications in the field of aerial photography,agricultural operation,rescue,small shipments and so on.Its high-quality operations need the improvement of maneuverability and control precision.This thesis focuses on the research of attitude estimation and control problem on large angle maneuvering targeted at the UAVs system.The precise and low delay attitude estimation is the key to improve the maneuverability and reliability of UAVs.In the attitude estimation problem of this thesis,the multi-sensor correction process is firstly carried out to acquire the correction parameters.Then,the Extended Kalman Filter(EKF)is further designed to achieve the multi-sensor fusion and attitude estimation.In order to relieve the communication burden of the monitoring from ground station,an event-based filter is developed to reduce the required communication bandwidth in remote manipulation and monitoring.With the specific filtering design,the reliable monitoring can be guaranteed even under complicated communication condition.Traditionally,UAV control methods employ Euler angles describing the attitude.The roll,pitch and yaw angle are controlled separately.However,the significant coupling between three axes is neglected,which result in degrading of the control performance.In order to overcome this problem,this thesis adopts the geometric control method,which calculates the attitude error in SO(3)directly to obtain the almost global stable control law.The application of this method can avoid the singularity and coupling problem encountered when using Euler description.It theoretically allows the effective control capability expanded from the local stability near equilibrium point to almost global stability.Therefore,it can stabilize the vehicle under large angle maneuvers.Finally,in order to verify the effectiveness of the algorithms,a quadrotor prototype was designed,including flight control system and the corresponding state monitoring ground station.The experiments are conducted on the UAV prototype.During the design process,many solutions are considered in order to balance the tradeoff among performance,sensor precision,cost,reliability,etc.The algorithms are optimized on the software level,achieving high effectiveness and reliability of flight control functions.The results verifies the validity of event-based EKF algorithm and geometric control algorithm.