The Design Of Multi-sensor Attitude Determination System And Research Of Attitude Algorithm

The traditional attitude measurement system is a kind of equipment which uses the high precision inertial device to acquire the yaw angle,pitch angle and roll angle of the motion carrier.It has played an important role in the aerospace,industrial control and other high-end fields.With the emergence of MEMS inertial sensors and magnetic sensors,it is possible to design a new multi-sensor attitude measurement system.The system has a very wide application prospect in civil and consumer electronics fields with the characteristics of low power consumption,small volume,low cost,high reliability,moderate precision and high integration.In this thesis,the attitude measurement system uses a 6-axis MEMS inertial device MPU6050,a magnetic sensor HMC5883 L and a barometer BMP180 as the sensitive elements.At the same time,this system also uses STM32F103T8U6 as the micro controller to collect the data of angular rate,acceleration,geomagnetic field,temperature and altitude of the carrier and executes data fusion algorithm to calculate the attitude information.Finally,the system outputs its attitude angle through the serial port.This thesis mainly studies the principle of attitude measurement based on strapdown inertial guidance technique,analyzes the quaternion representation of carrier rotation in three-dimensional space.The attitude matrix represented by quaternion and euler angle respectively and the updating equation of quaternion are derived in detail.The thesis also puts emphasis on description of the complementary filtering algorithm and the steepest descent algorithm used in this system.In addition,the parameters of each sensor are calibrated and the original output data is processed by filtering.Finally,the system's static output stability testing,dynamic tracking testing,attitude angle accuracy testing and the drift testing of gyroscope are performed.The results show that both algorithms can avoid angle drift and effectively compensate the output error of the attitude angle.In the static condition,the standard deviation of attitude angle of the complementary filtering algorithm and the steepest descent algorithm is better than 0.09° and 0.07° respectively.The latter can suppress the system noise more effectively and improve the stability of attitude calculation.