Investigation of Filtering Algorithm for Noise Reduction in Displacement Sensing Signal

With the emergence of new sensing technologies, soft substrate-based displacement sensing has gained popularity in the last decade. Nevertheless, performance degradation due to noise from unexpected vibrations of the substrate can affect the sensors accuracy and lead to poor measurements. Few studies have focused on noise reduction to improve the accuracy of the sensing data. Based on mathematical modeling, this paper compares the two most representative filtering algorithms based on minimum mean square error (MMSE), namely recursive least squares (RLS) and Kalman algorithms for their predictive ability, accuracy, and reliability. Based on the modeling and comparison, we propose a Kalman filtering algorithm to suppress the vibration noise in the displacement-based dosage sensor. The filter-processed data from the sensors measurement mainly contains the actual resistance change upon mechanical displacement of the sensor substrate and the undesired noise from its vibration. With the filtering algorithm, the noise can be reduced from the sensors measurement, and an accurate correlation between resistance change and displacement of the sensor can be established. This would provide a straightforward sensing signal that reflects the displacement, and in turn, achieves precise dosage monitoring in drug delivery devices, which can be used as a test-bed application for the sensor.

Automated summary

This paper proposes a Kalman filtering algorithm to suppress vibration noise in displacement sensors and improve data accuracy. The filtering algorithm effectively reduces noise, establishes accurate correlation between resistance change and sensor displacement, enabling precise dosage monitoring.