Research On Strong Tracking EKF And Active Disturbance Rejection Sliding Mode Control For Quadrotor Aircraft

Four rotors unmanned aerial vehicle(UAV)attitude algorithm and attitude control is one of the basic problems in its navigation.With the development of MEMS technology,MEMS inertial sensor costs more and more low,the precision is higher and higher,growing performance greatly promoted the development of the four rotors unmanned aerial vehicle,and four-rotor aircraft can flexible movement can also be hovering in the air,causing the extensive research of many scholars,quadrotor UAV has been involved in People’s daily life,such as aerial photography,express transportation,pesticide spraying,rescue and so on.In this paper,the research on quadrotors is mainly divided into two aspects: Research on improved strongly tracking EKF in attitude calculation and integrated navigation,and design of improved Active Disturbance Rejection Sliding Mode Control(ADRSMC)in attitude control.Firstly,this paper introduces the attitude algorithm,attitude control,integrated navigation-related theoretical basis,in improving the strong tracking EKF in attitude algorithm,integrated navigation in the study,based on EKF algorithm in tracking state change relatively quickly,because of the influence of measurement noise leads to an algorithm to estimate is not accurate,put forward the modified strong tracking filter,to enhance the system tracking ability,through the orthogonality principle,the variance of the new information sequence at a different time is kept orthogonal,and the suboptimal fading factor satisfying the orthogonality principle is introduced into the one-step predictive covariance matrix,and the extended Kalman gain is increased to increase the weight of the system to the observed value,to strengthen the tracking ability of the system.At the same time the introduction of Sage-Husa noise adaptive estimator to solve because of the strong tracking filter cannot test the accuracy of a single parameter decision criteria,only depends on the combination of all model parameters,a more accurate real-time estimate of the noise mean and variance of accelerometer and make the more accurate observation,suppresses the accelerometer noise caused by the new interest rates increase,every increase in new information detected by the system is due to a state mutation.Secondly,in the design of improved active disturbance rejection sliding mode compound controller for attitude control,aiming at the problem that the Extended State observer ESO in ADRC cannot effectively compensate the disturbance beyond the observation range,based on the compound idea,The sliding mode surface and switching rate of sliding mode control is introduced into ADRC to deal with uncertain systems and large range uncertainty problems,enhance the anti-interference ability of the system and improve system stability,and the stability of the introduced sliding mode controller is proved.At the same time,aiming at the chattering problem caused by the inflection point of traditional nonlinear FAL function,a smooth and continuous nonlinear NFAL function was used to redesign ESO,and the convergence of the improved ESO was proved.Finally,the effectiveness of the improved strong tracking extended Kalman algorithm is verified by an offline simulation experiment,online physical experiment,and integrated navigation simulation experiment in attitude solution,and the improved ADRSMC has a better disturbance rejection effect than the traditional ADRC.