Temperature Hysteresis Calibration Method of MEMS Accelerometer.

Micro-electromechanical system (MEMS) sensors are widely used in various navigation applications because of their cost-effectiveness, low power consumption, and compact size. However, their performance is often degraded by temperature hysteresis, which arises from internal temperature gradients. This paper presents a calibration method that corrects temperature hysteresis without requiring any additional hardware or modifications to the existing MEMS sensor design. By analyzing the correlation between the external temperature change rate and hysteresis errors, a mathematical calibration model is derived. The method is experimentally validated on MEMS accelerometers, with results showing an up to 63% reduction in hysteresis errors. We further evaluate bias repeatability, scale factor repeatability, nonlinearity, and Allan variance to assess the broader impacts of the calibration. Although minor trade-offs in noise characteristics are observed, the overall hysteresis performance is substantially improved. The proposed approach offers a practical and efficient solution for enhancing MEMS sensor accuracy in dynamic thermal environments.