Initial calibration of MEMS inertial measurement unit

To decrease the cost of calibrating MEMS inertial measurement unit (IMU) , this paper proposes an approach to calibrating MEMS IMU by means of machine vision technique. Mathematical models are derived from the three accelerometers and three rate gyros, taking into account the sensor axis misalignments, scale factors, and biases. Driven by external force, the IMU flies randomly between two screens. The collimated laser beams mounted on the plate of IMU produce four indicative spots on the screens. The coordinates of the indicative spots in the world coordinate system are measured by using binocular stereo reconstruction algorithm, and then the angular rate vector and acceleration vector of the body frame with respect to the world frame are determined by using Euler algorithm. Given this kinematics information, the calibration parameters of the mathematical models are solved by using least squares algorithm. The experimental results show that the angular rate vector and acceleration vector of the body frame with respect to the world frame can be determined accurately.