Determining efficient temperature test points for IMU calibration

Abstract

MEMS inertial sensors (gyroscope and accelerometer) are widely used in many areas from automotive to space applications. Inertial sensors are critical components in navigation applications as they provide valuable measurements such as acceleration and rotation. The structure and material used in MEMS inertial sensors are affected from some environmental factors such as vibration, mechanical shock, humidity, electromagnetism, and temperature. The most restrictive environmental factor for MEMS inertial sensors is temperature. The characteristics of noise and deterministic errors change under different temperature conditions. Therefore, these drawbacks of temperature have to be degraded before using inertial sensors in navigation applications. Generally, calibration tests are performed under different temperature conditions to identify the behavior of deterministic errors in the operating temperature range. This process is time-consuming; hence, many IMU producers try to determine the minimum sufficient number of temperature conditions, which are used in the calibration tests. In this study, deterministic error parameters are estimated for two different temperature conditions in separate calibration processes. One of the calibration processes is performed at eleven different temperature points to determine temperature dependent behavior of the deterministic errors in the operating temperature range. Then, second calibration process is carried out at five different temperature points in the operating temperature. Moreover, calibration verification tests have been performed under three different temperature conditions to examine how deterministic error parameters are affected by decreasing temperature points from 11 to 5 points. 8 uncalibrated IMUs from same batch are taken into consideration and their deterministic error characteristics are investigated.