Research On The Technology Of Strapdown Attitude Measurement System Based On MEMS Sensors

This dissertation takes Micro Electro Mechanical Systems (MEMS) based inertial sensors as measure element to study key technologies of the micro strapdown inertial attitude measurement system.MEMS based inertial sensors have superiorities in some aspects, such as small size and low cost. However, the precision of these sensors is much lower than other inertial sensors. Compared with MEMS accelerometers, the development of MEMS gyros is so slow that the precision couldn't meet the needs of inertial attitude measurement system so far. From the point of practical implementation view, a low-cost inertial attitude measurement system has been designed, which are composed of MEMS based inertial sensors, magnetometers, temperature sensors and aided by necessary signal sampling unit and micro processor.For the low resolution of sensors and large errors existed in the output signal, a sensor calibration method based on systemic calibration has been studied to ensure a satisfactory system output precision. After a discussion on the calibration varying rule, an effective means to compensate the constant part of the sensor error has been carried out.Due to the large power and wide frequency band, the stochastic noise in the output signal of MEMS gyros is difficult to eliminate. Therefore, on the basis of the principle study of Wavelet Analysis Methods, a wavelet de-noise technique is implied to the gyro output signal and examined by the practical experiments in order to specify appropriate wavelet function and according parameters.The paper makes use of the measurements of the accelerators and magnetometers to complete the system initial alignment, and use Kalman filter for data integration of the gyros, accelerometers and magnetometers to get an optimal estimation of vehicle attitude. The simulation shows that the designed strapdown inertial attitude measurement system has a good dynamic performance and could meet the design requirement.