کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

Estimating Position in Oscillating Motions By Development of a Physics-Informed Kalman Filter

نویسندگان :

M.Mohammadreza Ommatmohammadi¹ Morteza H Sadeghi² MohammadReza Sayyed Noorani³

1- دانشگاه تبریز 2- دانشگاه تبریز 3- دانشگاه تبریز

کلمات کلیدی :

Kalman filter،position estimation،Fourier expansion،drift reduction،tremor signal،IMU sensor

چکیده :

Position estimation using Inertial Measurement Units (IMUs) is crucial in various fields but is susceptible to drift error, especially in oscillating motions. This paper introduces a novel physics-informed Kalman Filter (KF) that explicitly incorporates the periodicity of motion as a physical constraint to mitigate drift. The method's accuracy was first verified using simulated oscillatory signals. Subsequently, it was successfully applied to estimate hand tremor in an experimental case study, demonstrating its practical value for developing tremor suppression devices. The results show superior accuracy and consistency without additional computational cost, confirming the framework's effectiveness and its potential for broader application in vibration analysis and other periodic motion tracking scenarios. Furthermore, the estimated kinematics provided by this approach pave the way for its application in inverse dynamics problems, where it can be utilized to approximate excitation forces, particularly when direct measurement of all kinematic variables is infeasible.