Characterization of error sources in an inertial unit using Allan variance and nonlinear parameter adjustment

Abstract

Locating in space is essential for devices that perform navigation. currently inertial measurement units have been developed to allow navigation, however, this type of sensors present different errors in measurements. They have different errors that appear and accumulate when measurements advance. For this reason, the characterization of the error sources presented in sensor BNO055 is performed; data is acquired using a Raspberry Pi and Allan's variance method was used for data analysis. A parameter fit is made in the error equation using the Levenberg-Marquardt numerical method a method commonly used in experimental physics.