The Algorithm Design And Implementation For Inertial Integrated Navigation System Based On MEMS Multi-sensor Data Fusion

With the continuous development and improvement of modern navigation system，on the one hand，the navigation system is widely used in the aerospace，marinetransportation，unmanned aircraft，missile guidance and control of industrial robots，automobile auto/driver assistance systems as well as the growing military and civilianareas. On the other hand， the modern electronic technology， especially thesemiconductor manufacture，as well as the rapid development of human spaceflightand the defense industry，requires one of the modern navigation systems with higherprecision，higher real-time and more miniaturization.As the emergence of the third-generation inertial navigation sensor material-MEMS (Micro Electro Mechanical Systems)， as well as the development andimprovement of the SINS theory，the miniaturized navigation system can be achieved；The high-performance microcontrollers based on ARM Cortex M4core with DSPfunctions offers a appropriate processor platform for the modern navigation systemwith more miniaturization，low power consumption and high intelligence.This thesis firstly focuses on the research of MEMS sensor error compensation forthe inertial navigation system，then proposes a kind of complementary filter andKalman filter based on multi-sensor data fusion theory to improve the accuracy ofnavigation sensors.The inertial navigation attitude solution based on the theory ofquaternions operator is introduced，after further modeling on the flight carrier，proposea kind of carrier flight path data generation algorithm，and then gives its softwaresimulation and analysis in the Matlab/Simulink. Then，for integrated GPS/SINSnavigation， consideration of both the accuracy of the algorithm and its easilyembedded implementation，propose the forecast residual vector-adaptive factor basedadaptive Kalman filter algorithm，verify and analysis it in Matlab to determine itsfeasibility. Finally，take implementation of this algorithm on the STM32F4basedembedded hardware platforms withthe10dimensions MEMS sensor data fusionalgorithm to verify the reliability and timeliness，the result turns out to meet the marketrequirements of a low-cost navigation system with miniaturization and high real-time.