Abstrato

Method for in-field user calibration of MEMS accelerometer using hybrid genetic algorithm

Sok Hun Kim*, Song Bong Jong, Gwang Jo Jong, Yu Song Choe

The Inertial Measurement Unit (IMU) using MEMS sensor contains 3-axis accelerometer, 3-axis gyroscope, 3-axis magnetometer, thermometer, etc. in a single microchip, but the installing axes are not ideally perpendicular to each other and calibration is needed. This paper describes the approach to the static calibration of an accelerometer without using any mechanical equipment on the basis of the fact that the norm of MEMS accelerometer outputs measured in the static position is ideally equal to 1. By using genetic algorithm, we verified the initial values of scale factors and zero bias ones. Taking these as the initial values of the Sequence Quadratic Programming (SQP), we found the optimal solution. We proved the effectiveness of the calibration using the measurements of MEMS accelerometer in the static position. The experimental result shows that the static calibration approach using the estimated Hybrid Genetic Algorithm (HGA) is better than the others.