Design of an Embedded Icosahedron Mechatronics for Robust Iterative IMU Calibration

Applications of low-cost Microelectromechanical Systems Inertial Measurement Units (MEMS-IMU) have broadened over the past few years. Currently, IMUs are implemented in a wide variety of consumer electronics such as smartphones, smartwatches, and AR/VR devices to detect human poses and real-time orientation of unmanned aerial vehicles. However, in order to precisely monitor objects poses, calibration of the IMU is crucial. In most cases, calibration procedures are conducted under a clean and perturbation free environment. However, such ideal environments and conditions may not always be available. Moreover, expensive and heavy calibration systems are usually required and unaffordable for general users. As a result, in this paper, a low cost, lightweight, and portable IMU calibration embedded icosahedron platform is developed. Since a clean calibration environment cannot be guaranteed, a calibration algorithm, with the consideration of external perturbations, is presented. An iterative weighted Levenberg-Marquardt algorithm (IWLMA) is proposed to cope with perturbations. The highly integrated hardware/software co-design provides user-friendly operations for IMU calibration. Finally, simulations, as well as experiments, are presented to demonstrate the effectiveness and robustness of the proposed system.