Research on Data-driven MEMS Accelerometer Self-detection and Self-correction Technology
A micro-electro-mechanical system(MEMS)accelerometer is miniature integrated system,which is widely used to measure carrier acceleration in various industrial and domestic applications.However,MEMS devices are susceptible to faults due to internal and external factors during operation.If the fault can not be detected and the fault data can not be corrected in time,the sys-tem will not be able to accurately perceive the external environment,resulting in control deviations.Hence,It is of great significance for the timely detection and correction of MEMS accelerometer faults to improve the robustness of the system measurement data accu-racy,and control stability.Existing detection and calibration methods often rely on establishing the accelerometer physical model or constructing redundant sensor networks to achieve the self-detection and self-correction of accelerometer,but these methods have the shortages of complexities in modeling,additional errors,or high hardware resource requirements.To avoid inaccuracies introduced by modeling and reduce the demand of the algorithm on hardware resources,a lightweight,data-driven self-testing and self-calibration al-gorithm for the MEMS accelerometer is proposed based on the notion of proximal sensor computation.Test results demonstrate that the algorithm achieves a detection rate of 90％for four types of faults:shock,bias,signal loss,and constant output.The average ab-solute error between the calibrated data and the normal data is less than 0.15 g,with the ability to process the accelerometer response data within 2.54 ms.
万方数据
维普
