This paper presents error modelling and error analysis of microelectromechnical systems (MEMS) inertial measurement unit (IMU) for a low-cost strapdown inertial navigation system (INS). The INS consists of IMU and navigation processor. The IMU provides acceleration and angular rate of the vehicle in all the three axes. In this paper, errors that affect the MEMS IMU, which is of low cost and less volume, are stochastically modelled and analysed using Allan variance. Wavelet decomposition has been introduced to remove the high frequency noise that affects the sensors to obtain the original values of angular rates and accelerations with less noise. This increases the accuracy of the strapdown INS. The results show the effect of errors in the output of sensors, easy interpretation of random errors by Allan variance, the increase in the accuracy when wavelet decomposition is used for denoising inertial sensor raw data. Allan variance of length T INTRODUCTIONThe inertial navigation system (INS) calculates velocity and position by integration of the total acceleration of the aircraft and integration of the resultant velocity, respectively. The strapdown INS eliminates most of the mechanical complexities of the gimbaled INS by having the sensors attached rigidly to the body of the aircraft, for the benefits of lower cost, reduced size, greater reliability, etc. The strapdown INS consists of inertial measurement unit (IMU) and navigation processor which computes position, velocity, and attitude based on navigation algorithm. The IMU consists of three accelerometers and three gyroscopes that provide accelerations along three axes and angular rates about three axes. The cost of strapdown INS is mainly determined by the IMU.The cost and the volume of these inertial sensors can be drastically reduced using microelectromechnical systems technology. The fabrication processes of MEMS sensors make these very sensitive to the changes in the surrounding environmental conditions like temperature, pressure, electric, and magnetic fields, etc. These changes cause the output

No abstract

No abstract

No abstract

No abstract