Noise Reduction of IMU Output Signals by Wavelet Transform with Allen Variance

In recent years, the advancement of inertial navigation systems (INS) with the aim of increasing road safety and underwater vehicles localization has dramatically enhanced. One of the most important parts of an inertial navigation system is the Inertial Measuring Unit (IMU), which is made up of accelerometers and gyroscope sensors. IMU output signals usually have different noises and errors. Neglect of these two problematic issues could lead to a sharp decline in the performance of a navigation system. Identifying, modeling and removing noises and errors are the most important challenges in an inertial navigation system. In this paper, a method for modeling the accelerometers and gyroscope sensors will be presented. This modeling, with 6 degree-of-freedom (DOF) for 3DM-GX25 inertial setup, is one of the efforts made in this study. Due to the dramatic growth of noise reduction methods by wavelet transformation in an inertial navigation system, the method is explored in detail in this paper. Finally, the performance of Allan variance and on removing noise and signal errors will be investigated. In this study, the effect of wavelet transformation denoising on the performance of a subsurface navigation signal system will be examined by ALVAR software. Simulations show that the use of wavelet transformation in 3DM-GX25 intrusion data for subsurface vehicles does not necessarily lead to improving navigation performance.