Research On Several Improved Control Algorithms Of Quadrotor Unmanned Aerial Vehicle Flight Attitude

Quadrotor Unmanned Aerial Vehicle(QUAV)is a kind of unmanned aerial vehicle,which can realize autonomous flight by remote control.It is an underactuated system with four driving DC motors which produces the differential torque to achieve pitch and roll motion and produces the reverse-torque to realize yaw motion.The flight trajectory can be controlled by the remote controller.The attitude of QUAV is easy to be disturbed by the factors such as the external air flow during the flight,which leads to the instability of the flight attitude and impacts the quality of the preset flight mission.Because of the characteristics of multi-variable,strong coupling,nonlinear and so on,the design of the attitude control system appears very important.Aiming at the traditional PID control parameters tuning requires skilled and sufficient time and the adjusted parameters fail to adaptively adjust external changes,it is difficult to use a fixed set of parameters to finish the control target mission.For this purpose,this thesis proposes several improved algorithms as follows:1.Firstly,this thesis puts forward a method of combination between elitist ant colony system(EACS)algorithm and PID to control the flight attitude of QUAV.Secondly,in order to overcome the defect that the pheromone is too large or too small in the path of EACS algorithm,an improved elitist ant colony system(IEACS)algorithm is presented.Compared with EACS algorithm,IEACS algorithm can get the optimal solution in a shorter time.Finally,these two optimization algorithms are combined with PID control respectively and then to control the flight attitude of QUAV.2.As the back-propagation neural network(BPNN)in the training process may not converge due to the oscillation,this thesis uses back-propagation neural network with inertia term(BPNNI)to solve this problem and improved the inertia coefficient,and then combined with PID control.A control method of combination between BPNNI algorithm and PID control is developed,and realizing the flight attitude control based on this control method.3.It can realize PID control parameters self-tuning to use ant colony algorithm(ACA)to optimize the PID control parameters.But the information will mislead the future cycle for the first path created by the ant colony is mainly the distance information.And best-worst ant system(BWAS)algorithm enhances the optimal solution and weakens the worst solution,which makes the increasing gap of pheromone and improves the optimization efficiency.Therefore,the control method of combination between BWAS algorithm and PID control is developed to control the flight attitude of QUAV.4.Based on the research of IEACS algorithm,active disturbance rejection control(ADRC)algorithm and PID control algorithm,the combination of these three algorithms named compound control(CC)is proposed by this thesis,which is used to control the flight attitude of QUAV.The simulation results show that the above four kinds of improved control algorithms contribute to control the flight attitude of QUAV.Comparing to the traditional PID control algorithm,the above improved algorithms can better adapt to the change of the system parameters itself and the influence of the external air flow and has a better performance of anti-disturbance and robustness.Thus,these improved algorithms have good practical values.