Design Of Quadrotor Attitude Measurement System Based On Error Quaternion UKF Algorithm

The quadrotor attitude measurement system is an important part of the quadrotor.For the problem that the MEMS attitude measurement system has low measurement accuracy and is greatly affected by external interference,this paper designs and implements a low-cost MEMS attitude measurement system,focusing on the attitude measurement system.The data fusion problem in the attitude measurement system is mainly studied.The system state equation and measurement equation are constructed by using the error quaternion method,and the UKF algorithm is used to realize data fusion to improve the attitude measurement accuracy.Considering the high dynamic environment of the attitude measurement system,the mechanical vibration is large,the external noise interference has a great influence on the sensor,the measurement noise has unknown and large transformation characteristics.An adaptive filtering algorithm is proposed,which adaptively adjusts the noise covariance matrix to ensure the smooth flight of quadrotor under high dynamic conditions.Finally,the overall function of the attitude measurement system is realized by using embedded system design,and its performance is verified by experiments.Firstly,it analyzes and summarizes the development of attitude measurement system at home and abroad as well as data fusion and attitude solution methods,and considers the actual cost and other factors to determine the overall scheme of the attitude measurement system.In this paper,low-cost MEMS gyroscopes,accelerometers and magnetometers are used to form the main sensors of the attitude measurement system.The data fusion and attitude calculation methods are combined to realize the measurement of the four-rotor attitude.Combined with data fusion and attitude calculation method,the attitude measurement of four rotors is realized.Then,the process of attitude calculation is analyzed,and the quaternion method is used for attitude calculation to achieve the full attitude of the attitude measurement system.The error source of MEMS sensor is analyzed to facilitate the establishment of a more accurate sensor model.Based on the quaternion,the sensor's attitude calculation method is analyzed to establish a reasonable sensor attitude measurement model.Then,the performance of the UKF filtering algorithm and the design implementation process are analyzed to determine the data fusion method of this paper.The state equation and measurement equation are constructed by using error quaternion method,which further reduces the state dimension and response speed of the system.For high dynamic flight,the measurement noise is easily affected by the disturbance.The adaptive error quaternion UKF filter algorithm is proposed to improve the measurement accuracy of the quadrotor attitude measurement system under high dynamics.Finally,the hardware of the attitude measurement system is designed and manufactured.At the same time,the function of the attitude measurement system is realized with the software module.Then the sensor data are pre-processed and corrected,and the performance of the attitude measurement system is verified with the quadrotor platform.The simulation and actual experimental results show that the attitude measurement system can reduce the influence of noise variation and improve the measurement accuracy of the quadrotor attitude measurement system under high dynamics.It has certain practical reference value for the design of the quadrotor attitude measurement system.