• CN: 11-2187/TH
  • ISSN: 0577-6686

Journal of Mechanical Engineering ›› 2026, Vol. 62 ›› Issue (10): 62-72.doi: 10.3901/JME.260491

Previous Articles    

Bias Stability Optimization Technique for Silicon Micro-resonant Accelerometer Based on Interval Temperature Compensation and Data Fusion

LI Ting1, ZHOU Wu1, LI He2,3, ZHANG Binglei2,3, MA Qingshan2,3, LI Ang2,3, MA Zhongrui1, WEI Zhaoli1, RAN Longqi1   

  1. 1. School of Mechanical and Electrical Engineering, University of Electronic Science and Technology of China, Chengdu 611731;
    2. Microsystem and Terahertz Research Center, China Academy of Engineering Physics, Chengdu 610299;
    3. Institute of Electronic Engineering, China Academy of Engineering Physics, Mianyang 621900
  • Received:2025-05-25 Revised:2025-11-19 Published:2026-07-29

Abstract: To address the deterioration of zero bias stability in silicon micro-resonant accelerometers under temperature varia tions, a collaborative optimization technique combining interval temperature compensation and data fusion is proposed. A dual double-ended tuning fork(DETF) temperature response simulation model is established to reveal the zero-point drift mechanism caused by thermal effects. An innovative interval temperature compensation algorithm is developed, employing sliding windows for real-time fitting of temperature coefficients and dynamically weighted compensation to eliminate temperature response differences between dual DETFs. Kalman filter-based data fusion technology is integrated to suppress residual noise, forming a synergistic optimization mechanism. Experimental validation is conducted using a measurement prototype built on the Zynq 7020 platform and a stress-isolated silicon micro-resonant accelerometer framework. Full-temperature testing under cyclic temperature variations ( -40-60 ℃) demonstrates that the collaborative approach achieves 95.19% improvement in zero bias stability compared to uncompensated conditions, confirming effectiveness in complex thermal environments. The method outperforms single-strategy approaches: 62.88% better than standalone temperature compensation and 99.9% better than pure data fusion. New insights are provided for optimizing MEMS accelerometer performance under significant temperature fluctuations.

Key words: MEMS, silicon micro-resonant accelerometer, temperature compensation, data fusion, zero-bias stability, FPGA

CLC Number: