Design And Research On Attitude And Heading Reference System Based On MEMS Inertial Sensor

With the rapid development of the MEMS(micro-electromechanical systems) technology, the AHRS(attitude and heading reference system) based on the MEMS interial sensor can be applied to the civil area in a wide range. In the inertial navigation technology based on the MEMS inertial sensor, the precision of the sensor is the major factor that determines the precision of the navigation system. Moreover, the temperature error of the gyroscope is the main factor that affects its precision. Therefore, establishing the temperature error model of the gyroscope is very important in order to improve the precision of the AHRS. This paper studies an AHRS that based on MEMS intertial sensor, then uses three methods to establish the tempertrue error model of the MEMS gyroscope, and lastly makes a research on the method to simulate the floating point arithmetic on fixed-point DSP.At beginging it introduces the basic theory and operation principle of the AHRS, including the definition of the common coordinate system, the attitude representation, the coordinate transformation and the basic equations of the inertial navigation system. Preliminary explain the constitution and working principle of the strapdown attitude and heading reference system. The working principle and characteristic of the MEMS inertial sensor were introduced also. Secondly, this paper gives the overall design scheme of the AHRS based on the MEMS inertial sensor. This system consists of the navigation data acquisition module, microprocessor and its peripheral circuit module and power supply module. The navigation data acquisition module includes the MEMS gyroscope ADXRS620 and MEMS accelerometer ADXL204 produced by Analog Devices, the magnetometer HMC1053 produced byHoneywell. The analog signal from these sensors will be converted to digital signal using the ADS1158 produced by TI before being sended to the DSP. The DSP in this system is the ADSP-BF506F produced by Analog Devices. 5V power supply is used in this system. There are some power conversion modules that convert the 5V power into different power that the system needs.Then, three methods including least squares modeling, BP nerual network and wavelet packet analysis, which can be used to establish the static output model and the scale factor model were introduced. Through the simulation result, we can conclude that the BP nerual is the best method to the application that need very high precision.Lastly, one method to simulate the floating-point arithmetic on the fixed-point DSP is proposed. Through the simulation result, we can know that this method can improve the accuracy of the data processing without affecting the real-time performance of the algorithm.