Application Of Multi-Sensors Data Fusion In A New Type Of Unmanned Rotorcraft

In this thesis,the multi-sensor data fusion technology was studied for the application requirements of new-type unmanned rotorcraft,and the attitude filter and integrated navigation system were mainly designed.In this thesis,the output model of the sensors used in the subject was studied firstly,and the online calibration software was designed to reduce the influence of the sensors' zero offset error and linear error on the accuracy of the sensors.In view of the high real-time requirements of engineering,in this thesis the state equation and observation equation of Extended Kalman Filter(EKF)were established by attitude four-element method,participating in time update with angular rate gyro,and participating in measurement update with accelerometer and magnetic heading meter.Based on the principle of feedback correction method,in this thesis the state variables of the integrated navigation system were selected,and the part of estimation of the attitude angle error in the traditional integrated navigation system was reduced,which simplified the state equation of the integrated navigation system and reduced the burden on the navigation computer.In this thesis,the structure of the Kalman Filter was improved and applied to attitude filtering and integrated navigation.The improved Kalman Filter would check the measurement sensor of the system and adjust the structure of the Kalman Filter to different degrees according to different degrees of interference.Since long-term large disturbances could force the system to only update the time,it might bring a large integral error.Therefore,after the disturbance disappears,the improved structure of the Kalman Filter will use the principle of strong tracking Kalman Filter to quickly recover the system.In addition,the Kalman Filter with improved structure will use the filter divergence criterion to perform health check on the filter,and if necessary,use strong tracking filter for divergence suppression to compensate for the defect of extended Kalman Filter and indirect method integrated navigation.Aiming at various problems that may exist in the application of attitude filtering and integrated navigation,in this thesis a Kalman Filter with improved structure was proposed.Through the semi-physical simulation and contrast experiments,it is verified that the improved Kalman Filter has higher filter accuracy,self-healing ability and stability.