微機電慣性測量儀的成本低廉、體積小且重量極輕，但因精度不高限制了應用的領域。率定及阿倫變方分析是常見用來提升或驗證慣性測量儀性能及精度的方法，兩者可分別檢驗系統及隨機誤差。台灣的學術領域對這方面的相關文獻並不多，業界又普遍使用進口的慣性元件，而使用者並沒有能力自行檢驗或調校。本研究的目的就是針對一款低階且未經任何誤差補償的MEMS IMU，發展一套一階段率定程序輔以阿倫變方分析，提升其精度，並應用在車載平台。

The Inertial Measuring Unit (IMU) has been used in many applications in our life. However, the accuracy of IMU limited where they can be used. The Micro Electro Mechanical Systems (MEMS) IMU is low cost but also low accuracy. The common ways to improve the accuracy of IMU are calibration and Allan Variance. The calibration can determine the systematic errors like bias, scale factor and non-orthogonality from accelerometers and gyros. The Allan variance can analyze the random errors include Angle Random walk (ARW), Velocity Random walk (VRW) and so on. The results of Allan variance can determine the parameters of those random errors with Power Spectral Density (PSD). There are only few researches about calibration and error analysis in Taiwan. The Taiwanese civil users have no ability to test and correct the IMU used in Mobile Mapping System (MMS) and other applications when some problems happened. The MEMS IMU we used didn’t do any error compensations and usually used in smartphone. The purpose is to develop one step calibration scheme and also the analysis of Allan variance for MEMS IMU. The field tests are used to verify the performance of IMU after calibration.