The Design Of MEMS INS/GPS Integrated Navigation System

The quartz tuning fork gyroscope and quartz vibrating beam accelerometer is newinertial devices by MEMS technology. In foreign countries, they have been widely used inautomotive safety, tactical missiles, smart bombs and other fields. In this thesis, based onLinde’s quartz tuning fork gyroscope and quartz vibrating beam accelerometer, developed aintegrated navigation system combined by GPS.FPGA, the system’s main chip, builted a SOPC embedded navigation computer.System’s main algorithms are the strapdown inertial navigation algorithm and the EKFalgorithm. Ultimately the basic functions of the integrated navigation system was realized.The test results showed that the horizontal attitude accuracy is0.4°, and long-term work. Inaddition, in dynamic tracking, the heading attitude accuracy is less than0.5°.The thesis mainly includes:(1) Build a SOPC-based embedded systems, privided a hardware experimental platformfor the study of navigation algorithms and error compensation. Through the analysis of tuningfork gyroscope, vibration beam accelerometer, GPS and geomagnetic sensor’s principles, thehardware system was designed, including the circuit board design on the structure of MIMU,electrical connections, the design of frequency measurement module by verilog language, thebuild of SOPC kernel and the bottom-driven design.(2) Based on the hardware experimental platform, analysed the inertial devices andMIMU’s models and errors. To the random error of the collected data, designed the digitalfilter and Allan analysis. Calibration of inertial device, then utilized the MIMU static errormodel designed the MIMU calibration experiments, and compensated its error last. Tested thedevice’s nonlinearity. Finally, tested its repeatability, and obtained the bias, bias stability andbias repeatability.(3) A brief introduction of the strapdown navigation and initial alignment principles,simulated the strapdown algorithm using simulated data. Combined with the accuracies of theinertial devices, specified the EKF algorithm, which ignore the Coriolis acceleration and theEarth’s rotation. Through the simulation using actual sampling data, the attitude error is±0.05°, the speed error is±0.2m/s and the position error is-4m. It’s the guidance for actualembedded software design.(4) Finally, designed the static and dynamic experiments, analysed the results, andproposed recommendations for improvement of the system.