Research Of Attitude Determination And Control For Quadcopter

In recent years,research on the theory and application of unmanned aerial vehicles,such as fixed wing and multicopter,has become more and more popular.Many universities and research institutions have invested a large amount of funds for relevant research.Compared with other aircraft,the quadcopter is relatively simple and the price is relatively low.But it is a complex and unstable underactuated system with six degrees of freedom and four actuators.Because of the nonlinear coupling between freedom and actuator,the stability control of the quadcopter becomes more difficult.As a result,it has become the preferred research platform for mathematical modeling,simulation and control algorithm verification in many universities.In this paper,the following work is done to solve the problems of the quadcopter in its attitude control,sensor calibration and attitude determination module design of strapdown inertial navigation system:1.The inertial measurement unit composed of the MEMS sensors used in the strapdown inertial navigation system is characterized by low accuracy,high noise,and easy to cause integral drift.Firstly,the inertial sensor and magnetometer error model are established and the calibration algorithm is optimized based on the original algorithm.Then the sensor data is filtered.Experimental results show that the algorithm can significantly reduce the sensor's measurement error and improve the sensor's anti-jamming capability.2.An improved adaptive unscented Kalman Filter for attitude determination algorithm(IAUKF)is proposed.By combining the unscented Kalman filter with the adaptive Sage-Husa algorithm,the statistical characteristics of the measured noise of the observation equation are updated online.Only three attitude angles are updated per iteration.This algorithm can effectively solve the drift and noise problems of MEMS gyroscopes and reduce the impact of acceleration on accelerometers and improve the anti-jamming capability of the attitude updating algorithm.And The truncation error of the extended Kalman filter updating algorithm in linearization is avoided.Its static RMS error and dynamic RMS error are less than the older.This algorithm is especially suitable for the design of strapdown inertial navigation systems for small and medium sized quadcopter with high requirements for attitude determination accuracy.3.For the problem that the PID controller is easy to oversaturate and overshoot in the attitude control process of the quadcopter,in order to further improve the control effect,the structure of the attitude loop PID controller is further optimized on the basis of the traditional double closed loop PID controller.The PD-ADRC cascade controller is designed by using the auto disturbance rejection controller instead of the PID controller of the angular velocity loop.Simulation and tests in MATLAB/SIMULINK have achieved the desired results.There is almost no overshoot in attitude and angular velocity tracking curves.This control strategy effectively improves the stability of the attitude control.Finally,The IAUKF attitude updating algorithm and the PD-ADRC attitude control algorithm are further verified and applied on the autuopilot which is designed and manufactured by myself.The algorithm achieve a good control effect and achieve a stable control of the quadcopter.